CNN.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "CNN.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Xvpkc97aWAA8",
        "outputId": "ddcc3af4-8a13-42af-8e25-b29e306aa377"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Done\n"
          ]
        }
      ],
      "source": [
        "from zipfile import ZipFile\n",
        "file_name=\"signal.zip\"\n",
        "\n",
        "with ZipFile(file_name,'r') as zip:\n",
        "  zip.extractall()\n",
        "  print('Done')"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline\n",
        "import os\n",
        "import cv2\n",
        "import tensorflow as tf\n",
        "import numpy as np \n",
        "import pandas as pd \n",
        "import seaborn as sns\n",
        "import warnings\n",
        "warnings.filterwarnings('ignore')"
      ],
      "metadata": {
        "id": "G-Mb_G7IWBus"
      },
      "execution_count": 534,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "IMG_SIZE=50\n",
        "dataset='traffic'\n",
        "categories=['GuideSign','M1','M4','M5','M6','M7','P1','P10_50','P12','W1']\n",
        "           \n",
        "      "
      ],
      "metadata": {
        "id": "e8tMNjf_dafd"
      },
      "execution_count": 535,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Categories_Data\n",
        "training_data=[]\n",
        "\n",
        "for c in categories:\n",
        "    path=os.path.join(dataset,c)\n",
        "    class_num=categories.index(c)\n",
        "    for img in os.listdir(path):\n",
        "  \n",
        "            img_arrary=cv2.imread(os.path.join(path,img))\n",
        "            new_array=cv2.resize(img_arrary,(IMG_SIZE,IMG_SIZE))\n",
        "            training_data.append([new_array,class_num])\n",
        "\n"
      ],
      "metadata": {
        "id": "CE0DqTCMz5JL"
      },
      "execution_count": 536,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import random\n",
        "random.shuffle(training_data)"
      ],
      "metadata": {
        "id": "N3jRXODPdkPK"
      },
      "execution_count": 537,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "for sample in training_data[:5]:\n",
        "    print(sample[1])\n",
        "print(new_array)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "HdqbvzNBdkUt",
        "outputId": "2afdf3f6-bfc8-42ed-f201-a5bfc52f25c6"
      },
      "execution_count": 538,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1\n",
            "5\n",
            "3\n",
            "5\n",
            "2\n",
            "[[[136 130 123]\n",
            "  [138 131 122]\n",
            "  [141 133 120]\n",
            "  ...\n",
            "  [126 117 104]\n",
            "  [132 123 109]\n",
            "  [133 125 108]]\n",
            "\n",
            " [[132 131 123]\n",
            "  [134 132 122]\n",
            "  [137 134 120]\n",
            "  ...\n",
            "  [113 105  92]\n",
            "  [121 114  99]\n",
            "  [125 118 101]]\n",
            "\n",
            " [[125 130 121]\n",
            "  [128 132 121]\n",
            "  [131 133 118]\n",
            "  ...\n",
            "  [ 94  88  75]\n",
            "  [104  99  84]\n",
            "  [113 109  91]]\n",
            "\n",
            " ...\n",
            "\n",
            " [[ 55 103 105]\n",
            "  [ 32  83  85]\n",
            "  [  9  63  64]\n",
            "  ...\n",
            "  [ 16  62  69]\n",
            "  [ 31  66  67]\n",
            "  [ 52  71  59]]\n",
            "\n",
            " [[ 48  97  99]\n",
            "  [ 28  79  81]\n",
            "  [ 10  64  65]\n",
            "  ...\n",
            "  [ 26  70  75]\n",
            "  [ 37  70  69]\n",
            "  [ 52  70  57]]\n",
            "\n",
            " [[ 41  89  91]\n",
            "  [ 25  76  78]\n",
            "  [ 10  64  65]\n",
            "  ...\n",
            "  [ 34  76  79]\n",
            "  [ 41  72  69]\n",
            "  [ 52  69  54]]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "X=[]\n",
        "y=[]"
      ],
      "metadata": {
        "id": "h2QgonWkjbz6"
      },
      "execution_count": 539,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "for feature ,label in training_data: \n",
        "    X.append(feature)\n",
        "    y.append(label)\n",
        "plt.imshow(X[1])\n",
        "Q=X[0]/255\n",
        "Q\n",
        "print(len(y))\n",
        "print(len(X))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 303
        },
        "id": "dTr638iBjcBA",
        "outputId": "59a9cfe6-6765-45ce-def1-a5df8dce3179"
      },
      "execution_count": 540,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "6348\n",
            "6348\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "X=np.array(X)\n",
        "y=np.array(y)\n"
      ],
      "metadata": {
        "id": "xqhXA5QgjcKt"
      },
      "execution_count": 541,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "plt.figure(figsize=(17,5))\n",
        "g = sns.countplot(y, palette=\"icefire\")\n",
        "plt.title(\"Number of digit classes\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 354
        },
        "id": "uKk3ZVQRj56n",
        "outputId": "ddd466fa-ec60-4424-d133-003c6128126f"
      },
      "execution_count": 542,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Number of digit classes')"
            ]
          },
          "metadata": {},
          "execution_count": 542
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1224x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "X_train,X_test,y_train,y_test=train_test_split(X,y,test_size=0.2,shuffle=True,random_state=2)\n",
        "X_Train,X_val,y_Train,y_val=train_test_split(X_train,y_train,test_size=0.2,random_state=2)"
      ],
      "metadata": {
        "id": "nBh9Jeopj_Hk"
      },
      "execution_count": 543,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(len(X_Train))\n",
        "print(len(X_val))\n",
        "print(len(X_test))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sMi-fS-zj_bY",
        "outputId": "9bcf71a4-3e52-4691-ad97-c24fc0371d98"
      },
      "execution_count": 544,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "4062\n",
            "1016\n",
            "1270\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "X_Train.shape\n",
        "y_Train[:5]\n",
        "X_Train.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "aJ1bQ4iBj_fg",
        "outputId": "322d3d4c-add7-4ff6-8dc2-39e0a6fa041d"
      },
      "execution_count": 545,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(4062, 50, 50, 3)"
            ]
          },
          "metadata": {},
          "execution_count": 545
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.matshow(X_Train[0])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 293
        },
        "id": "nM4ddObTkH3V",
        "outputId": "5f6bb773-291d-4d57-f0d3-e60a991b4218"
      },
      "execution_count": 546,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7fcb1aae1fd0>"
            ]
          },
          "metadata": {},
          "execution_count": 546
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 288x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Normalization\n",
        "\n",
        "X_Train = X_Train / 255.0\n",
        "X_test = X_test / 255.0\n"
      ],
      "metadata": {
        "id": "ljKC2NhXkMhg"
      },
      "execution_count": 547,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#CNN\n",
        "\n",
        "from tensorflow.keras import layers, models,optimizers\n",
        "from tensorflow.keras.models import Sequential\n",
        "cnn = Sequential([\n",
        "    layers.Conv2D(filters=8, kernel_size=(3, 3),padding='same',activation='relu', input_shape=(50, 50,3)),\n",
        "    layers.MaxPooling2D((2, 2)),\n",
        "    layers.Dropout(0.25),\n",
        "\n",
        "    layers.Conv2D(filters=16, kernel_size=(3, 3),padding='same',activation='relu'),\n",
        "    layers.MaxPooling2D((2, 2)),\n",
        "    layers.Dropout(0.25),\n",
        "\n",
        "  \n",
        "  \n",
        "\n",
        "    layers.Flatten(),\n",
        "    layers.Dense(256, activation='relu'),\n",
        "    layers.Dense(128, activation='relu'),\n",
        "    layers.Dropout(0.25),\n",
        "    layers.Dense(10, activation='softmax')\n",
        "    \n",
        "])\n",
        "cnn.summary()"
      ],
      "metadata": {
        "id": "aByWf01ckMnA",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "8b49b5c9-52ab-432d-bf96-fa298f637958"
      },
      "execution_count": 548,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_54\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv2d_152 (Conv2D)         (None, 50, 50, 8)         224       \n",
            "                                                                 \n",
            " max_pooling2d_152 (MaxPooli  (None, 25, 25, 8)        0         \n",
            " ng2D)                                                           \n",
            "                                                                 \n",
            " dropout_207 (Dropout)       (None, 25, 25, 8)         0         \n",
            "                                                                 \n",
            " conv2d_153 (Conv2D)         (None, 25, 25, 16)        1168      \n",
            "                                                                 \n",
            " max_pooling2d_153 (MaxPooli  (None, 12, 12, 16)       0         \n",
            " ng2D)                                                           \n",
            "                                                                 \n",
            " dropout_208 (Dropout)       (None, 12, 12, 16)        0         \n",
            "                                                                 \n",
            " flatten_54 (Flatten)        (None, 2304)              0         \n",
            "                                                                 \n",
            " dense_162 (Dense)           (None, 256)               590080    \n",
            "                                                                 \n",
            " dense_163 (Dense)           (None, 128)               32896     \n",
            "                                                                 \n",
            " dropout_209 (Dropout)       (None, 128)               0         \n",
            "                                                                 \n",
            " dense_164 (Dense)           (None, 10)                1290      \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 625,658\n",
            "Trainable params: 625,658\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "import time\n",
        "\n",
        "\"\"\"\n",
        "class TimeHistory(tf.keras.callbacks.Callback):\n",
        "    def on_train_begin(self, logs={}):\n",
        "        self.times = []\n",
        "\n",
        "    def on_epoch_begin(self, epoch, logs={}):\n",
        "        self.epoch_time_start = time.time()\n",
        "\n",
        "    def on_epoch_end(self, epoch, logs={}):\n",
        "        self.times.append(time.time() - self.epoch_time_start)\n",
        "        \"\"\"\n",
        "#time_callback1 = TimeHistory()\n",
        "tcnn=time.time()\n"
      ],
      "metadata": {
        "id": "KjkWPjt3d4jp"
      },
      "execution_count": 549,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "adm=optimizers.Adam(learning_rate=0.001)\n",
        "cnn.compile(optimizer=adm,\n",
        "          loss='sparse_categorical_crossentropy',\n",
        "      metrics=['accuracy'])"
      ],
      "metadata": {
        "id": "s71Mf14tjqy9"
      },
      "execution_count": 550,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "history1=cnn.fit(X_Train,y_Train,validation_data = (X_val,y_val),verbose='auto',epochs=35,batch_size=32)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "B1G5n3MhkMso",
        "outputId": "ee87f2c3-b763-46ad-f550-a7d35f025a2b"
      },
      "execution_count": 551,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/35\n",
            "127/127 [==============================] - 10s 78ms/step - loss: 1.0959 - accuracy: 0.6657 - val_loss: 45.0759 - val_accuracy: 0.8209\n",
            "Epoch 2/35\n",
            "127/127 [==============================] - 5s 43ms/step - loss: 0.5603 - accuracy: 0.8301 - val_loss: 53.1840 - val_accuracy: 0.8406\n",
            "Epoch 3/35\n",
            "127/127 [==============================] - 5s 42ms/step - loss: 0.4480 - accuracy: 0.8646 - val_loss: 39.7817 - val_accuracy: 0.8760\n",
            "Epoch 4/35\n",
            "127/127 [==============================] - 5s 43ms/step - loss: 0.3785 - accuracy: 0.8808 - val_loss: 31.1736 - val_accuracy: 0.8829\n",
            "Epoch 5/35\n",
            "127/127 [==============================] - 8s 62ms/step - loss: 0.3106 - accuracy: 0.9057 - val_loss: 30.1217 - val_accuracy: 0.9085\n",
            "Epoch 6/35\n",
            "127/127 [==============================] - 5s 42ms/step - loss: 0.2701 - accuracy: 0.9183 - val_loss: 28.9546 - val_accuracy: 0.9026\n",
            "Epoch 7/35\n",
            "127/127 [==============================] - 5s 42ms/step - loss: 0.2251 - accuracy: 0.9338 - val_loss: 31.7906 - val_accuracy: 0.9026\n",
            "Epoch 8/35\n",
            "127/127 [==============================] - 5s 43ms/step - loss: 0.2092 - accuracy: 0.9372 - val_loss: 24.5477 - val_accuracy: 0.9055\n",
            "Epoch 9/35\n",
            "127/127 [==============================] - 5s 42ms/step - loss: 0.1813 - accuracy: 0.9468 - val_loss: 49.5429 - val_accuracy: 0.8868\n",
            "Epoch 10/35\n",
            "127/127 [==============================] - 5s 42ms/step - loss: 0.1745 - accuracy: 0.9473 - val_loss: 30.6405 - val_accuracy: 0.9065\n",
            "Epoch 11/35\n",
            "127/127 [==============================] - 5s 42ms/step - loss: 0.1423 - accuracy: 0.9572 - val_loss: 30.8647 - val_accuracy: 0.9075\n",
            "Epoch 12/35\n",
            "127/127 [==============================] - 5s 42ms/step - loss: 0.1325 - accuracy: 0.9618 - val_loss: 34.6857 - val_accuracy: 0.8976\n",
            "Epoch 13/35\n",
            "127/127 [==============================] - 5s 43ms/step - loss: 0.1261 - accuracy: 0.9631 - val_loss: 25.7094 - val_accuracy: 0.8957\n",
            "Epoch 14/35\n",
            "127/127 [==============================] - 5s 43ms/step - loss: 0.1048 - accuracy: 0.9643 - val_loss: 29.7560 - val_accuracy: 0.9173\n",
            "Epoch 15/35\n",
            "127/127 [==============================] - 5s 43ms/step - loss: 0.0986 - accuracy: 0.9695 - val_loss: 40.1558 - val_accuracy: 0.8730\n",
            "Epoch 16/35\n",
            "127/127 [==============================] - 5s 42ms/step - loss: 0.0947 - accuracy: 0.9727 - val_loss: 24.3858 - val_accuracy: 0.9094\n",
            "Epoch 17/35\n",
            "127/127 [==============================] - 5s 42ms/step - loss: 0.0849 - accuracy: 0.9746 - val_loss: 44.7300 - val_accuracy: 0.8652\n",
            "Epoch 18/35\n",
            "127/127 [==============================] - 5s 42ms/step - loss: 0.0764 - accuracy: 0.9749 - val_loss: 64.2214 - val_accuracy: 0.8425\n",
            "Epoch 19/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.0636 - accuracy: 0.9796 - val_loss: 58.2241 - val_accuracy: 0.8839\n",
            "Epoch 20/35\n",
            "127/127 [==============================] - 5s 42ms/step - loss: 0.0674 - accuracy: 0.9791 - val_loss: 36.5770 - val_accuracy: 0.9183\n",
            "Epoch 21/35\n",
            "127/127 [==============================] - 5s 42ms/step - loss: 0.0618 - accuracy: 0.9813 - val_loss: 41.0726 - val_accuracy: 0.8976\n",
            "Epoch 22/35\n",
            "127/127 [==============================] - 6s 46ms/step - loss: 0.0651 - accuracy: 0.9810 - val_loss: 41.0409 - val_accuracy: 0.8858\n",
            "Epoch 23/35\n",
            "127/127 [==============================] - 5s 42ms/step - loss: 0.0492 - accuracy: 0.9825 - val_loss: 109.3655 - val_accuracy: 0.8278\n",
            "Epoch 24/35\n",
            "127/127 [==============================] - 5s 43ms/step - loss: 0.0571 - accuracy: 0.9838 - val_loss: 76.1896 - val_accuracy: 0.8553\n",
            "Epoch 25/35\n",
            "127/127 [==============================] - 5s 43ms/step - loss: 0.0463 - accuracy: 0.9867 - val_loss: 61.6550 - val_accuracy: 0.8740\n",
            "Epoch 26/35\n",
            "127/127 [==============================] - 5s 43ms/step - loss: 0.0410 - accuracy: 0.9850 - val_loss: 67.1881 - val_accuracy: 0.8287\n",
            "Epoch 27/35\n",
            "127/127 [==============================] - 5s 43ms/step - loss: 0.0436 - accuracy: 0.9879 - val_loss: 57.6870 - val_accuracy: 0.8947\n",
            "Epoch 28/35\n",
            "127/127 [==============================] - 5s 42ms/step - loss: 0.0457 - accuracy: 0.9840 - val_loss: 45.4437 - val_accuracy: 0.8996\n",
            "Epoch 29/35\n",
            "127/127 [==============================] - 5s 43ms/step - loss: 0.0412 - accuracy: 0.9852 - val_loss: 46.6374 - val_accuracy: 0.8967\n",
            "Epoch 30/35\n",
            "127/127 [==============================] - 5s 42ms/step - loss: 0.0423 - accuracy: 0.9874 - val_loss: 31.7686 - val_accuracy: 0.9035\n",
            "Epoch 31/35\n",
            "127/127 [==============================] - 5s 42ms/step - loss: 0.0337 - accuracy: 0.9899 - val_loss: 54.0141 - val_accuracy: 0.8957\n",
            "Epoch 32/35\n",
            "127/127 [==============================] - 5s 43ms/step - loss: 0.0218 - accuracy: 0.9931 - val_loss: 51.3170 - val_accuracy: 0.8927\n",
            "Epoch 33/35\n",
            "127/127 [==============================] - 5s 42ms/step - loss: 0.0292 - accuracy: 0.9902 - val_loss: 70.2080 - val_accuracy: 0.8878\n",
            "Epoch 34/35\n",
            "127/127 [==============================] - 5s 42ms/step - loss: 0.0323 - accuracy: 0.9884 - val_loss: 129.0526 - val_accuracy: 0.7530\n",
            "Epoch 35/35\n",
            "127/127 [==============================] - 5s 42ms/step - loss: 0.0340 - accuracy: 0.9879 - val_loss: 114.4846 - val_accuracy: 0.8337\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from numpy.lib.function_base import average\n",
        "time1=(time.time()-tcnn)\n",
        "print(time1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "w_Xql68_fD53",
        "outputId": "4884ba49-1ab1-48d5-b71c-db98dac696ae"
      },
      "execution_count": 552,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "201.97470712661743\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Evaluate Training_Date\n",
        "\n",
        "trainD1=cnn.evaluate(X_Train,y_Train)\n",
        "trainD1"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hTsfBxnH6cA2",
        "outputId": "3700052d-252e-497e-b626-d4d8a27bd37f"
      },
      "execution_count": 553,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "127/127 [==============================] - 2s 16ms/step - loss: 0.0210 - accuracy: 0.9938\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.021014634519815445, 0.9938454031944275]"
            ]
          },
          "metadata": {},
          "execution_count": 553
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_acc1=trainD1[1]"
      ],
      "metadata": {
        "id": "UFNPBVG6UKfC"
      },
      "execution_count": 554,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#Evaluate Testing_Date\n",
        "\n",
        "testD1=cnn.evaluate(X_test,y_test)\n",
        "testD1"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5vqIFqMWlN94",
        "outputId": "7eed6131-2411-416c-b722-a0af3aba829a"
      },
      "execution_count": 555,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "40/40 [==============================] - 1s 16ms/step - loss: 0.2742 - accuracy: 0.9543\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.2742493748664856, 0.9543306827545166]"
            ]
          },
          "metadata": {},
          "execution_count": 555
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_acc1=testD1[1]"
      ],
      "metadata": {
        "id": "XPe8ZktXg_Ur"
      },
      "execution_count": 556,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#Evaluate Validation_Date\n",
        "\n",
        "cnn.evaluate(X_val,y_val)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "O-Fe6_-GFP2E",
        "outputId": "01f6517d-f099-4c22-eba1-7c512c7250c0"
      },
      "execution_count": 557,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "32/32 [==============================] - 1s 16ms/step - loss: 114.4846 - accuracy: 0.8337\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[114.48458862304688, 0.8336614370346069]"
            ]
          },
          "metadata": {},
          "execution_count": 557
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Dilated_CNN Model\n",
        "\n",
        "di = Sequential([\n",
        "    layers.Conv2D(filters=8, kernel_size=(3, 3),padding='same',activation='relu',dilation_rate=2, input_shape=(50, 50,3)),\n",
        "    layers.MaxPooling2D((2, 2)),\n",
        "     layers.Dropout(0.25),\n",
        "     layers.Conv2D(filters=16, kernel_size=(3, 3),padding='same',activation='relu'),\n",
        "    layers.MaxPooling2D((2, 2)),\n",
        "     layers.Dropout(0.25),\n",
        "    \n",
        "\n",
        "\n",
        "   layers.Flatten(),\n",
        "    layers.Dense(256, activation='relu'),\n",
        "    layers.Dense(128, activation='relu'),\n",
        "     layers.Dropout(0.25),\n",
        "    layers.Dense(10, activation='softmax')\n",
        "    \n",
        "])\n",
        "di.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LlWkK5pfKZHE",
        "outputId": "6e0a9013-0543-46f2-d0dc-c904d8ba060a"
      },
      "execution_count": 558,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_55\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv2d_154 (Conv2D)         (None, 50, 50, 8)         224       \n",
            "                                                                 \n",
            " max_pooling2d_154 (MaxPooli  (None, 25, 25, 8)        0         \n",
            " ng2D)                                                           \n",
            "                                                                 \n",
            " dropout_210 (Dropout)       (None, 25, 25, 8)         0         \n",
            "                                                                 \n",
            " conv2d_155 (Conv2D)         (None, 25, 25, 16)        1168      \n",
            "                                                                 \n",
            " max_pooling2d_155 (MaxPooli  (None, 12, 12, 16)       0         \n",
            " ng2D)                                                           \n",
            "                                                                 \n",
            " dropout_211 (Dropout)       (None, 12, 12, 16)        0         \n",
            "                                                                 \n",
            " flatten_55 (Flatten)        (None, 2304)              0         \n",
            "                                                                 \n",
            " dense_165 (Dense)           (None, 256)               590080    \n",
            "                                                                 \n",
            " dense_166 (Dense)           (None, 128)               32896     \n",
            "                                                                 \n",
            " dropout_212 (Dropout)       (None, 128)               0         \n",
            "                                                                 \n",
            " dense_167 (Dense)           (None, 10)                1290      \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 625,658\n",
            "Trainable params: 625,658\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#time_callback2 = TimeHistory()\n",
        "tdi=time.time()"
      ],
      "metadata": {
        "id": "Inocdskehk-e"
      },
      "execution_count": 559,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "adm=optimizers.Adam(learning_rate=0.001)\n",
        "di.compile(optimizer=adm,\n",
        "          loss='sparse_categorical_crossentropy',\n",
        "      metrics=['accuracy'])"
      ],
      "metadata": {
        "id": "LGXjpTA2KZKC"
      },
      "execution_count": 560,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "history2=di.fit(X_Train,y_Train,validation_data = (X_val,y_val),verbose='auto',epochs=35,batch_size=32)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "a7FCb0ztKZP3",
        "outputId": "09179988-0871-4758-8322-ea99e3a8b576"
      },
      "execution_count": 561,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/35\n",
            "127/127 [==============================] - 6s 46ms/step - loss: 1.0344 - accuracy: 0.6709 - val_loss: 45.6834 - val_accuracy: 0.8278\n",
            "Epoch 2/35\n",
            "127/127 [==============================] - 6s 45ms/step - loss: 0.5750 - accuracy: 0.8257 - val_loss: 32.7602 - val_accuracy: 0.8642\n",
            "Epoch 3/35\n",
            "127/127 [==============================] - 6s 45ms/step - loss: 0.4491 - accuracy: 0.8644 - val_loss: 38.8006 - val_accuracy: 0.8770\n",
            "Epoch 4/35\n",
            "127/127 [==============================] - 6s 45ms/step - loss: 0.3860 - accuracy: 0.8808 - val_loss: 34.7047 - val_accuracy: 0.8632\n",
            "Epoch 5/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.3305 - accuracy: 0.9015 - val_loss: 27.1021 - val_accuracy: 0.9055\n",
            "Epoch 6/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.2767 - accuracy: 0.9121 - val_loss: 36.7304 - val_accuracy: 0.8868\n",
            "Epoch 7/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.2322 - accuracy: 0.9291 - val_loss: 30.4611 - val_accuracy: 0.9104\n",
            "Epoch 8/35\n",
            "127/127 [==============================] - 6s 45ms/step - loss: 0.2199 - accuracy: 0.9357 - val_loss: 30.7578 - val_accuracy: 0.9075\n",
            "Epoch 9/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.1815 - accuracy: 0.9429 - val_loss: 35.1707 - val_accuracy: 0.9213\n",
            "Epoch 10/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.1638 - accuracy: 0.9493 - val_loss: 29.2366 - val_accuracy: 0.9104\n",
            "Epoch 11/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.1472 - accuracy: 0.9586 - val_loss: 29.6755 - val_accuracy: 0.9144\n",
            "Epoch 12/35\n",
            "127/127 [==============================] - 6s 45ms/step - loss: 0.1238 - accuracy: 0.9623 - val_loss: 30.1098 - val_accuracy: 0.9213\n",
            "Epoch 13/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.1089 - accuracy: 0.9695 - val_loss: 36.9982 - val_accuracy: 0.9272\n",
            "Epoch 14/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.1014 - accuracy: 0.9685 - val_loss: 29.5513 - val_accuracy: 0.9193\n",
            "Epoch 15/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.0976 - accuracy: 0.9705 - val_loss: 36.8950 - val_accuracy: 0.9154\n",
            "Epoch 16/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.0903 - accuracy: 0.9729 - val_loss: 36.9862 - val_accuracy: 0.9341\n",
            "Epoch 17/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.0794 - accuracy: 0.9791 - val_loss: 27.4163 - val_accuracy: 0.9183\n",
            "Epoch 18/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.0611 - accuracy: 0.9818 - val_loss: 35.6303 - val_accuracy: 0.9134\n",
            "Epoch 19/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.0655 - accuracy: 0.9806 - val_loss: 37.5143 - val_accuracy: 0.8957\n",
            "Epoch 20/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.0564 - accuracy: 0.9823 - val_loss: 46.3191 - val_accuracy: 0.9213\n",
            "Epoch 21/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.0554 - accuracy: 0.9803 - val_loss: 42.0425 - val_accuracy: 0.9262\n",
            "Epoch 22/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.0456 - accuracy: 0.9855 - val_loss: 26.6670 - val_accuracy: 0.9419\n",
            "Epoch 23/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.0518 - accuracy: 0.9855 - val_loss: 34.4250 - val_accuracy: 0.9429\n",
            "Epoch 24/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.0389 - accuracy: 0.9894 - val_loss: 58.6104 - val_accuracy: 0.8927\n",
            "Epoch 25/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.0314 - accuracy: 0.9919 - val_loss: 57.4013 - val_accuracy: 0.9094\n",
            "Epoch 26/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.0449 - accuracy: 0.9874 - val_loss: 44.0435 - val_accuracy: 0.9400\n",
            "Epoch 27/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.0317 - accuracy: 0.9899 - val_loss: 42.6536 - val_accuracy: 0.9390\n",
            "Epoch 28/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.0323 - accuracy: 0.9906 - val_loss: 44.9966 - val_accuracy: 0.9380\n",
            "Epoch 29/35\n",
            "127/127 [==============================] - 6s 45ms/step - loss: 0.0436 - accuracy: 0.9852 - val_loss: 30.6709 - val_accuracy: 0.9390\n",
            "Epoch 30/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.0281 - accuracy: 0.9909 - val_loss: 44.6156 - val_accuracy: 0.9498\n",
            "Epoch 31/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.0460 - accuracy: 0.9877 - val_loss: 56.2136 - val_accuracy: 0.9429\n",
            "Epoch 32/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.0305 - accuracy: 0.9894 - val_loss: 42.2168 - val_accuracy: 0.9360\n",
            "Epoch 33/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.0233 - accuracy: 0.9924 - val_loss: 57.7947 - val_accuracy: 0.9085\n",
            "Epoch 34/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.0411 - accuracy: 0.9860 - val_loss: 35.1916 - val_accuracy: 0.9429\n",
            "Epoch 35/35\n",
            "127/127 [==============================] - 6s 44ms/step - loss: 0.0272 - accuracy: 0.9897 - val_loss: 34.1548 - val_accuracy: 0.9400\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "time2=(time.time()-tdi)\n",
        "print(time2)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ohjS2-P9jwYK",
        "outputId": "413d5563-5550-4043-a26d-696bbef9aa43"
      },
      "execution_count": 562,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "206.64966583251953\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Evaluate Training_Date\n",
        "\n",
        "trainD2=di.evaluate(X_Train,y_Train)\n",
        "trainD2"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "RrbpDi52KZUp",
        "outputId": "e07a14cb-02cc-48db-860e-5721d0d37c13"
      },
      "execution_count": 563,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "127/127 [==============================] - 2s 17ms/step - loss: 0.0122 - accuracy: 0.9966\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.012183218263089657, 0.9965534210205078]"
            ]
          },
          "metadata": {},
          "execution_count": 563
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_acc2=trainD2[1]"
      ],
      "metadata": {
        "id": "_Xq1rr_sioWI"
      },
      "execution_count": 564,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#Evaluate Testing_Date\n",
        "\n",
        "testD2=di.evaluate(X_test,y_test)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "upwjdYFrKZcF",
        "outputId": "af9280a2-7d5b-4ba5-e02e-41d2e3f689d6"
      },
      "execution_count": 565,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "40/40 [==============================] - 1s 17ms/step - loss: 0.2588 - accuracy: 0.9488\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_acc2=testD2[1]"
      ],
      "metadata": {
        "id": "0qRseGyIi2c1"
      },
      "execution_count": 566,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#Evaluate Validation_Date\n",
        "\n",
        "di.evaluate(X_val,y_val)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "d0jql3qMFb0M",
        "outputId": "6db30b9c-1f57-4ae9-ef21-a0d817858f02"
      },
      "execution_count": 567,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "32/32 [==============================] - 1s 16ms/step - loss: 34.1548 - accuracy: 0.9400\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[34.154815673828125, 0.9399606585502625]"
            ]
          },
          "metadata": {},
          "execution_count": 567
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#HDC Model\n",
        "from sklearn.neural_network import MLPClassifier\n",
        "\n",
        "hdc = Sequential([\n",
        "    layers.Conv2D(filters=8, kernel_size=(3, 3),padding='same',activation='relu',dilation_rate=2, input_shape=(50,50,3)),\n",
        "    layers.MaxPooling2D((2, 2)),\n",
        "    layers.Dropout(0.25),\n",
        "\n",
        "    layers.Conv2D(filters=16, kernel_size=(3, 3),padding='same',activation='relu',dilation_rate=3),\n",
        "    layers.MaxPooling2D((2, 2)),\n",
        "    layers.Dropout(0.25),\n",
        "\n",
        " \n",
        "\n",
        "\n",
        "   layers.Flatten(),\n",
        "    layers.Dense(256, activation='relu'),\n",
        "    layers.Dense(128, activation='relu'),\n",
        "    layers.Dropout(0.25),\n",
        "    layers.Dense(10, activation='softmax')\n",
        "    \n",
        "])\n",
        "hdc.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "j7hU7qWoBnUZ",
        "outputId": "795fdf1d-5fa6-4b9e-edd4-4d56582a1c1e"
      },
      "execution_count": 495,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_50\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv2d_144 (Conv2D)         (None, 50, 50, 8)         224       \n",
            "                                                                 \n",
            " max_pooling2d_144 (MaxPooli  (None, 25, 25, 8)        0         \n",
            " ng2D)                                                           \n",
            "                                                                 \n",
            " dropout_195 (Dropout)       (None, 25, 25, 8)         0         \n",
            "                                                                 \n",
            " conv2d_145 (Conv2D)         (None, 25, 25, 16)        1168      \n",
            "                                                                 \n",
            " max_pooling2d_145 (MaxPooli  (None, 12, 12, 16)       0         \n",
            " ng2D)                                                           \n",
            "                                                                 \n",
            " dropout_196 (Dropout)       (None, 12, 12, 16)        0         \n",
            "                                                                 \n",
            " flatten_50 (Flatten)        (None, 2304)              0         \n",
            "                                                                 \n",
            " dense_150 (Dense)           (None, 256)               590080    \n",
            "                                                                 \n",
            " dense_151 (Dense)           (None, 128)               32896     \n",
            "                                                                 \n",
            " dropout_197 (Dropout)       (None, 128)               0         \n",
            "                                                                 \n",
            " dense_152 (Dense)           (None, 10)                1290      \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 625,658\n",
            "Trainable params: 625,658\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#time_callback3 = TimeHistory()\n",
        "thdc=time.time()"
      ],
      "metadata": {
        "id": "SAUQHbpriG-u"
      },
      "execution_count": 496,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "adm=optimizers.Adam(learning_rate=0.001)\n",
        "hdc.compile(optimizer=adm,\n",
        "          loss='sparse_categorical_crossentropy',\n",
        "      metrics=['accuracy'])\n",
        "     "
      ],
      "metadata": {
        "id": "rURoP8H_BncD"
      },
      "execution_count": 497,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "history3=hdc.fit(X_Train,y_Train,validation_data = (X_val,y_val),verbose='auto',epochs=35,batch_size=32)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6RD5r64EBni0",
        "outputId": "3a994b0d-484c-4134-942f-14c41b791d5f"
      },
      "execution_count": 498,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/35\n",
            "127/127 [==============================] - 7s 48ms/step - loss: 1.0489 - accuracy: 0.6733 - val_loss: 37.2115 - val_accuracy: 0.7766\n",
            "Epoch 2/35\n",
            "127/127 [==============================] - 6s 48ms/step - loss: 0.5681 - accuracy: 0.8203 - val_loss: 28.1798 - val_accuracy: 0.8514\n",
            "Epoch 3/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.4370 - accuracy: 0.8708 - val_loss: 22.9622 - val_accuracy: 0.8898\n",
            "Epoch 4/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.3759 - accuracy: 0.8880 - val_loss: 70.6662 - val_accuracy: 0.7421\n",
            "Epoch 5/35\n",
            "127/127 [==============================] - 6s 48ms/step - loss: 0.3232 - accuracy: 0.9028 - val_loss: 34.8482 - val_accuracy: 0.8287\n",
            "Epoch 6/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.2700 - accuracy: 0.9195 - val_loss: 29.7350 - val_accuracy: 0.8839\n",
            "Epoch 7/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.2351 - accuracy: 0.9286 - val_loss: 30.3363 - val_accuracy: 0.8789\n",
            "Epoch 8/35\n",
            "127/127 [==============================] - 6s 48ms/step - loss: 0.2068 - accuracy: 0.9350 - val_loss: 25.0497 - val_accuracy: 0.9035\n",
            "Epoch 9/35\n",
            "127/127 [==============================] - 7s 58ms/step - loss: 0.1780 - accuracy: 0.9485 - val_loss: 28.5599 - val_accuracy: 0.8612\n",
            "Epoch 10/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.1659 - accuracy: 0.9490 - val_loss: 19.9562 - val_accuracy: 0.9114\n",
            "Epoch 11/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.1483 - accuracy: 0.9535 - val_loss: 30.0817 - val_accuracy: 0.8917\n",
            "Epoch 12/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.1166 - accuracy: 0.9648 - val_loss: 28.5024 - val_accuracy: 0.8927\n",
            "Epoch 13/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.1192 - accuracy: 0.9626 - val_loss: 24.8342 - val_accuracy: 0.9104\n",
            "Epoch 14/35\n",
            "127/127 [==============================] - 6s 46ms/step - loss: 0.1026 - accuracy: 0.9668 - val_loss: 32.5893 - val_accuracy: 0.8986\n",
            "Epoch 15/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.0956 - accuracy: 0.9690 - val_loss: 27.6529 - val_accuracy: 0.8917\n",
            "Epoch 16/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.0884 - accuracy: 0.9732 - val_loss: 33.8615 - val_accuracy: 0.8829\n",
            "Epoch 17/35\n",
            "127/127 [==============================] - 6s 49ms/step - loss: 0.0815 - accuracy: 0.9749 - val_loss: 39.2703 - val_accuracy: 0.8937\n",
            "Epoch 18/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.0764 - accuracy: 0.9756 - val_loss: 43.5121 - val_accuracy: 0.8780\n",
            "Epoch 19/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.0650 - accuracy: 0.9808 - val_loss: 33.7223 - val_accuracy: 0.9183\n",
            "Epoch 20/35\n",
            "127/127 [==============================] - 6s 46ms/step - loss: 0.0521 - accuracy: 0.9840 - val_loss: 40.9545 - val_accuracy: 0.9134\n",
            "Epoch 21/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.0610 - accuracy: 0.9806 - val_loss: 39.6589 - val_accuracy: 0.9035\n",
            "Epoch 22/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.0508 - accuracy: 0.9840 - val_loss: 48.8746 - val_accuracy: 0.8878\n",
            "Epoch 23/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.0501 - accuracy: 0.9850 - val_loss: 40.3244 - val_accuracy: 0.8967\n",
            "Epoch 24/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.0533 - accuracy: 0.9830 - val_loss: 36.7256 - val_accuracy: 0.9134\n",
            "Epoch 25/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.0414 - accuracy: 0.9877 - val_loss: 59.2864 - val_accuracy: 0.8720\n",
            "Epoch 26/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.0444 - accuracy: 0.9857 - val_loss: 44.9118 - val_accuracy: 0.9085\n",
            "Epoch 27/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.0434 - accuracy: 0.9872 - val_loss: 39.0719 - val_accuracy: 0.9114\n",
            "Epoch 28/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.0486 - accuracy: 0.9830 - val_loss: 56.7400 - val_accuracy: 0.8425\n",
            "Epoch 29/35\n",
            "127/127 [==============================] - 6s 48ms/step - loss: 0.0381 - accuracy: 0.9884 - val_loss: 68.9629 - val_accuracy: 0.8504\n",
            "Epoch 30/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.0340 - accuracy: 0.9899 - val_loss: 53.6818 - val_accuracy: 0.8937\n",
            "Epoch 31/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.0398 - accuracy: 0.9872 - val_loss: 45.6688 - val_accuracy: 0.9154\n",
            "Epoch 32/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.0442 - accuracy: 0.9877 - val_loss: 88.0445 - val_accuracy: 0.8287\n",
            "Epoch 33/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.0362 - accuracy: 0.9874 - val_loss: 54.1342 - val_accuracy: 0.8819\n",
            "Epoch 34/35\n",
            "127/127 [==============================] - 6s 47ms/step - loss: 0.0320 - accuracy: 0.9911 - val_loss: 72.1861 - val_accuracy: 0.8681\n",
            "Epoch 35/35\n",
            "127/127 [==============================] - 6s 48ms/step - loss: 0.0314 - accuracy: 0.9894 - val_loss: 44.7980 - val_accuracy: 0.9075\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "time3=(time.time()-thdc)\n",
        "print(time3)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PKKdEewqj00r",
        "outputId": "72986925-6c30-42e8-fde6-627ba2de41fb"
      },
      "execution_count": 499,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "349.05693340301514\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Evaluate Training_Date\n",
        "\n",
        "trainD3=hdc.evaluate(X_Train,y_Train)\n",
        "trainD3"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_wcNZeUdBnp9",
        "outputId": "15c563dc-1b31-4a80-8432-683bc58db0ad"
      },
      "execution_count": 500,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "127/127 [==============================] - 4s 30ms/step - loss: 0.0163 - accuracy: 0.9963\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.016346564516425133, 0.9963072538375854]"
            ]
          },
          "metadata": {},
          "execution_count": 500
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_acc3=trainD3[1]"
      ],
      "metadata": {
        "id": "_O7noQINjK8o"
      },
      "execution_count": 501,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#Evaluate Testing_Date\n",
        "\n",
        "testD3=hdc.evaluate(X_test,y_test)\n",
        "testD3"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VbjGXEv7Bn0H",
        "outputId": "373076ed-f180-4985-ee2b-816037be0d74"
      },
      "execution_count": 502,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "40/40 [==============================] - 1s 36ms/step - loss: 0.1352 - accuracy: 0.9717\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.1351684182882309, 0.9716535210609436]"
            ]
          },
          "metadata": {},
          "execution_count": 502
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_acc3=testD3[1]"
      ],
      "metadata": {
        "id": "x9D5iJ9GjVqk"
      },
      "execution_count": 503,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#Evaluate Validation_Date\n",
        "\n",
        "hdc.evaluate(X_val,y_val)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7vSQYEpQFeXB",
        "outputId": "eff0c6a5-3d63-4a81-c755-ba13f97e4984"
      },
      "execution_count": 504,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "32/32 [==============================] - 1s 17ms/step - loss: 44.7980 - accuracy: 0.9075\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[44.7979621887207, 0.9074802994728088]"
            ]
          },
          "metadata": {},
          "execution_count": 504
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "CNN_Predict"
      ],
      "metadata": {
        "id": "Y0JaxFC_8UKB"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#Predict data cnn\n",
        "\n",
        "y_pred_cnn = cnn.predict(X_test)\n",
        "y_pred_classes_cnn = [np.argmax(element) for element in y_pred_cnn]\n",
        "y_pred_classes_cnn[:5]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BqJ9uCKelQcB",
        "outputId": "1ac0453e-f805-431f-a0eb-98e293f9f8ad"
      },
      "execution_count": 568,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[2, 0, 0, 2, 0]"
            ]
          },
          "metadata": {},
          "execution_count": 568
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "predictions_cnn=cnn.predict(X_test)\n",
        "predictions_cnn"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "M_Mg0BtTlUXp",
        "outputId": "0c72f2c3-9849-4c2e-fe05-da86a73192d5"
      },
      "execution_count": 569,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[ 0.0,  0.0,  1.0, ...,  0.0,  0.0,  0.0],\n",
              "       [ 1.0,  0.0,  0.0, ...,  0.0,  0.0,  0.0],\n",
              "       [ 1.0,  0.0,  0.0, ...,  0.0,  0.0,  0.0],\n",
              "       ...,\n",
              "       [ 0.0,  0.0,  1.0, ...,  0.0,  0.0,  0.0],\n",
              "       [ 0.7,  0.1,  0.1, ...,  0.0,  0.0,  0.0],\n",
              "       [ 0.0,  0.0,  1.0, ...,  0.0,  0.0,  0.0]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 569
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_predicted_labels_cnn = [np.argmax(i) for i in y_pred_cnn]"
      ],
      "metadata": {
        "id": "g6ETpnRXlZAV"
      },
      "execution_count": 570,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "score=tf.nn.softmax(predictions_cnn[0])\n",
        "score\n",
        "np.argmax(score)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nhcFSNxslcYD",
        "outputId": "a63c75cb-ba45-4e74-9c89-37146f32ea93"
      },
      "execution_count": 571,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "2"
            ]
          },
          "metadata": {},
          "execution_count": 571
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.imshow(X_test[0])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 285
        },
        "id": "ypKrmeyillox",
        "outputId": "cee059f5-3988-4791-d19e-a8ce69f3da88"
      },
      "execution_count": 572,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7fcb1b03f410>"
            ]
          },
          "metadata": {},
          "execution_count": 572
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "CNN_ConfutionMatrix\n"
      ],
      "metadata": {
        "id": "Fd2XZCi8Aaep"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#Confusion Matrix\n",
        "\n",
        "from sklearn.metrics import confusion_matrix, accuracy_score\n",
        "cm = tf.math.confusion_matrix(labels=y_test,predictions=y_predicted_labels_cnn)\n",
        "print(cm)\n",
        "accuracy_score(y_test, y_predicted_labels_cnn)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rQrCHL4Zlly7",
        "outputId": "83db3d1a-c166-43a3-ddd3-d05f078d35c7"
      },
      "execution_count": 573,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tf.Tensor(\n",
            "[[234   0   7   1   0   2   0   0   0   0]\n",
            " [  0  43   9   0   0   2   0   0   0   0]\n",
            " [ 12   4 648   0   1   3   0   0   0   0]\n",
            " [  0   0   3  38   1   0   0   0   0   0]\n",
            " [  4   0   0   2  24   0   0   0   0   0]\n",
            " [  1   2   0   0   0  95   0   0   0   0]\n",
            " [  1   0   1   0   0   0  65   0   0   0]\n",
            " [  0   0   0   0   0   0   0  24   0   0]\n",
            " [  1   0   1   0   0   0   0   0  16   0]\n",
            " [  0   0   0   0   0   0   0   0   0  25]], shape=(10, 10), dtype=int32)\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9543307086614173"
            ]
          },
          "metadata": {},
          "execution_count": 573
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "CNN_ROC"
      ],
      "metadata": {
        "id": "30Nqp-O4B-u0"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#ROC Model\n",
        "\n",
        "from sklearn.metrics import roc_curve, auc\n",
        "import matplotlib.pyplot as plt\n",
        "from collections import Counter\n",
        "\n",
        "y_pred1 = cnn.predict(X_test)\n",
        "y_pred1=[np.argmax(i) for i in y_pred1]\n",
        "\n",
        "plt.figure(figsize=(10,5))\n",
        "\n",
        "\n",
        "for i in range(10):\n",
        "    nn_fpr_keras, nn_tpr_keras, nn_thresholds_keras = roc_curve(y_test  , y_pred1, pos_label=i)\n",
        "    auc_keras = auc(nn_fpr_keras, nn_tpr_keras)\n",
        "    plt.plot(nn_fpr_keras, nn_tpr_keras,marker='.'.format(), label='class # %0.f' % i)\n",
        "    \n",
        "plt.legend(loc=\"lower right\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 338
        },
        "id": "nsDyBnWwll4E",
        "outputId": "a769c774-27fc-4b53-e9dc-1f65b22029ca"
      },
      "execution_count": 574,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7fcb1abadc10>"
            ]
          },
          "metadata": {},
          "execution_count": 574
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x360 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlMAAAEvCAYAAABhSUTPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOy9eXxc5Xn3/b3PmZE0Wm1Zkhd5N7uXkbwJh2JS3DqBLCR2yJsSUzsEAs4boOYpT9xCimOWt27d0jZNajAOrv34LY3LkhCTB963BOLEBAO2AQuIF2FbNots2ZLG1kgzc879/DGLRvuMZkY6I11fPmJmztxzzn3k0ZzfXNfvvi6ltUYQBEEQBEEYGMZQT0AQBEEQBCGbETElCIIgCIKQAiKmBEEQBEEQUkDElCAIgiAIQgqImBIEQRAEQUgBEVOCIAiCIAgp4BqqA5eVlempU6cO1eEFQRAEQRAS5q233jqjtS7v6bkhE1NTp07lzTffHKrDC4IgCIIgJIxS6nhvz0maTxAEQRAEIQVETAmCIAiCIKSAiClBEARBEIQUEDElCIIgCIKQAiKmBEEQBEEQUkDElCAIgiAIQgqImBIEQRAEQUiBfsWUUuqnSqkGpdTBXp5XSql/UUodUUq9o5Sam/5pCoIgCIIgOJNEinZuBf4V2NbL89cBF0d+aoB/i9wOKW82X2BP03k+M6qQ+SUFCT8Xz65Ddfz3x6dZMr6cL1wyPaNzfeHYKSY0n+Zz0yYzadKkjB0rFRL9faTjfAbrdy8IwvCidf9+Wve+we6LLmW3Ky+1z5D6vbT+93/R2pDL7nl/kvr+hIzw95se4dDofC4518q9d/z1kMxBaa37H6TUVOCXWutZPTz3GPCK1vo/Io//AHxWa/1xX/ucP3++zlQF9DebL/DV/UcIao0BXFGYR5HLBMAXsnjvfBs2dHsunsbWNg61h2KPx7W34sFO+1zblMEnOflE/xXKzjdTYBoYhkr7sVLBj8Enufmxx739PtJxPokeSxCEEY4GtI79aMtCB9rxu3JoGDUmNmxsawsey0py3zaEAmgL/O4u+zvfhMcK9fFiIXF03P/pdj9+Q8f28PVEGRbBHPgkrwIN5BDkzkP/nTFBpZR6S2s9v6fn0tFOphKoj3t8MrKtm5hSSn0H+A7A5MmT03DontnTdJ5gRCTaQEvIigmmlpAVuyx3fS6epkAAMEAp0Bq/MvDoDIgpwxV+g0SO0+5y47GCGEb3OQ0lfhXJCPfz+0jH+SR6LEEQshwd+Z/WkRvdSRyhNaAjNz091zNtObnhO9HPEAzy2i4MeJrd9me4yPP7B7y/waDTb6fH77Kq+7hehnUVMQBKqc6D4u+rnraDij5WXfbVdd/R0UqjjFDkx8IwQigzfF8ZIRSaM5ShUaAUIW1yaHTHF/HBZFB782mtHwceh3BkKlPH+cyoQgzCYinPUPzkiqmxdN6bzRf42oEjBG2Nu8tz8ew6VMftJ85iKQNTa/5x+tiMhHbfbL7A8v2HCVg2prZZeng/a5fd4LhUX/j3cQ5LqT5/H+k4n0SPJQjC0KJtG/vCBazmFmxfC1ZzC5avBbulBavF13lbcwuWz4fV0ozd4sPy+dD9CBLldmOUlGAWF2MWFWGUFGMWFWMUF2EWl2AWF2EUh7eZxUUEPv6ET9ev57czZ7Pulrtjn98bxxVx3aRLEjqnU0cPUX/kfSpKNMW7fkjT74r53ez5rLv1LyL7s1ljnKK6vI1gKBj5CRCyQgSDQYKhEJYdIhQKErIsrFAQK2QRskLYlo1lhbAtK+7HxrbDt9qy0baNtjTatsHSYGu0Hb5VtgYbTG3g0m5c2oVbu4n+54r+p9yYyo1LuTCjP0b0vhtTmZHbrs+54rbHjTEGLhVsbWFhESKErUJYWFjKwlYWIRXCyvFh5Z/BKmjE9pzD9pxF553DzjsHOec7v99CuditpQRbxhA6PwYdGMPRAhf/POMrhLSJC4tLzrUOeK6pMCzTfAB/8sYHtISsHsVSop6p+7b+L47k5PPn86vEM4V4pgRhOGK3tWG1dBFALS2dtsULoNj9lhZsn6/PCBFKYRQVoYoKUUUFUFSILixAF+ZjF3qwCzxYhR6sglyCBbmECnIJFuQQ8LgJ5LsJuCCkQwStICE7RNAOErQ77sdvG3OmgLFnSvC1HsU4Xsvrl86hbtQMio1j5OV+2G18/D7c5y1KGwwmfZRD5WkPAGNyJ1CWP5nT/hOcbfuIk5Mv5fiki5lSf5hJJw7FTtHAiAiOOHESL0yMrts6P2coE9NwYxidBY8r/j6R+5iYuLpEhRJHo7ENjTY12gRtEg6puBTKZaDcBsptYrpNjBwXZq4bM8eNmevGyA1vU26j09iO+3E/0W0uE60CtLV9RKv/OH5/PX7/Cfz+E7T56/G31WNZ8eJHkZs7Do9nMh7PJJQ9nuaPivn4D/l8esiDFSikdEIh06vLmVFdwZjKAk6e8/OTJ/+eMxNLMu6Z6ivNlw4x9QXge8D1hI3n/6K1XtjfPjMtpr66/zAAz1ZfPOB9PPnkkwB861vfSsucBEEQEuVAwwHe/PRN5o+dz8wxM2MioKsQCATbCPmaCTU3x0WGfNgtPvCdR5+/gPJdgPMXMM63Ypxvw7jQhutCANMfxLQ0KBOlIrYGZULc/VCOSTDPTcDjpj3PRXuui/Zck/YcI3zrVrTlKNpdijY3tLnCP+0usFwGBiYGCkMbmJgYWsVuDQxMHfe8Noi+otPjyG14H+FbF2YkCmNSEizk4tbJKBQazbHCj2lzB1EqfIzwrUIpA4XCsMBsszDaLVSbBaGIjUABGtxGLmNyJ0TSUprz7gtggKnN8Hy1iaENDFt1pK6SRREnOjqECD0JFFfX+2YPwiX+NWYP2www1ICFWG9orQmFmvD762n1H6fNX09rRDD5/Sdob/+E+EShYXjweCbFBJPHMxlP3iQ8nink5VXS/GmIo/sbOLr/NI0nw5Gp8slFzJhbzvSqckaP6xwAOfypjz999Df8603VfHHOhLSeW1dS8kwppf4D+CxQppQ6CTwAuAG01puAFwgLqSNAKyDKQxCEQUPrcAoEO/5+NDXSwzZNp9TJocZDfND4PpeNuoyLSi7qGBsZ1/E6jbY7toWsEMFQkFAkzRO0goRCkfSOHSIUCmFZ0R8rnN6xbWzLwrIsbDuc1rFtG4IWRtDGCNmYIY0R+bnCNgjYb/CufhPTNnFpA1OryAVdYWKQhxm+QCoDVA6oclBjIwJJgWGiTQNGKdToiGjKBKHITwpoCK8MMgClUAZgqMj98A+GgTKj98PjbDuITRAI+3IuMqZhFuTEdmqFggT8fgL+VgJtfuxQCDBQhpucfA85Hg85eR60bXP241PkGfkoOoTHqJJy3GMLehE28Y/NPqI0nUUQZvqFTaaw7RDt7R91E0xtkceW1Tkdl5NTjsczmdGja/B4pkTEUlgw5eSUdTpvrTVn6s/z7u8aqNu/n3OftIKC8dNLuOprFzG9qpziMk+vc2sNhBcWeNxD6zPuV0xprf+sn+c18H+nbUaCMMKJXuTDF++IIOi0jbgLfvRi331b/PZu23RnwdF5X5FtukNwxAuRnrbFi5PexvT/ui4CyOpyvnHzid/Wv4O2b0YBVzIF8HOGdwe0DwXkRH46trixtYnGhUajtQ3aiszZRtnhH2wbpa2I+brzGCLbNXY4NWPYkdSMjW0aWC4D3BpcJuQY4FaoHDcq142Rm4PKzcXIdWOaLpRhYJomhmFgmi5Ml4lhmBFRolAKMDuEC/Hbo0JGqYgIIO55FXkdHfsyIibkbq8LC6DO+1dx4mlg4qL9eAtnnngXHbJRLgP358r4uOkw9bXvcvL9g5w/2wiAp7iESZfPYuLM2Uy6fBZjJk5GGZ3FpXHofT79/Qfk/0GBZaMIUfr1KnKnFA9obtlCKOTD7z8REUkn4lJy9bS1n0LrjtWQSuXg8UzE45lMyai5XQTTJEyzbxO4tjWfHmvh6L4G6g6cpuVMG8pQTLh4FLM/O5Hp1eUUlOQmNG9/MEvElJAYbceaCdQ1kzOtmNzJg/9H99HhDzhZe5CJM2cx4eLL+hx76nfv0vz2SUq8E6m8anafY9tPtBD4sKXH80r0mO0nWgjUtZAzpZiciYW9Xky7RQK6XISj0YFO4qKTUIgbq+NEQux1ceIi0X3pLnOx4vfVwzkkLC56E0AJ/GMPJZFoQPgCGL1wEndR7OVi2tOFOfJjxF+Yjc4XZZRCmT1cmI2OyERP2zpe13lMx3HD57Dr2Au8cPwFbG1hhmwKLRejdR75IYP8kMIT0HgC4Ana5LZb5LZZ5LSFyPEHcbcGcLW247rQimoPRMRP5Me26KTyXC7MkmLM4pIO83RREUZph3naKC6OmKqLORL6mL8+8AjNOSGC+W4e+/wTVFVUDcE/uLP56ND71Ne+S3HFWGyvnwt/OEPdRwc49XcfAJBfMopJV8xm4hWzmXTFbEorJ/YbDZpwyeVMuORy2o+30P7zJ8gN7iF3ypLBOJ2MorVFe/unsfSbP14wtdUTDJ7rNN7tLsXjmURxiZexeV8MC6ZIWi43d2zSEU7bsvnoSDN1EQF1oTmAYSomXV7KvOumMs1bhqcwp/8ddSEqpvJyRExlPe3HWzjz2Dspf0NOlUrGo99o5BS/63dsCcVwqoVTL/Q/Nl3HdBqxb/w9/Ecv28PPaexweKfLf3a3cT1tDx83vL37Pnrbj93D6/vbTw/70j1v739fTld5hCM/VjjSo20LIiujsG20ZUW22xB337ZCLLAsjF7+dkOAL/KDYYSjGGY4VcPoXCjzoIwxnZ8zjEgqyoTofUPRfX26H1r80PIpnOp+7C+FvPhDfjwuDwfefpwD6fxlDQOC7e20nGnoZIAvGF3KpCtm8ydfWcKkmbMZPb5ywKm03CnF5I7dB2dOpmvKGceyWjuZvP3+evxt0fun0DoQG6uUi7y8CXjyJlNRcV3MtxSNLrlcRanPJ2Rz8g/nwgLq7TO0nQ/ichtMnjWG6VXlTJ1TRq4nNRnSli1pPqF/2uuaOwmp3ItGkTu9ZNCOf6L2HU68+3bscUHJKDzFPUfHcn25lKpxKKXQ2uas/pT2ovYexxYGSigKjoqZOn3uJs7nNAPgb2nhQnNTv8fsuo+WnHP4cpo6RILSkV9d9H50O50ECarjESpeaNDldXH76+d1A/WNOpPOF+ueLt1OR2sNoRA6FEIHg+hQEB2M3g/f0sO22HN2P4LPNFEuF8rtDt/m5BE0NJ8GzuDJK6Q4vxSPp7hjjNuFcoVvcbkGbjQWMsLZUyc7hJRSzLv+Bq65+dvp9SHZNjio5p/WmkCgobtg8h/H31ZPIHCm03iXqwiPZzKFhZdTXrY0zvg9mdzc8RgplDzojVDA4sR7Zzm6v4Fj7zQS8Idw55lMnV3GjOpyJs8cgzs3fb9TSfMNI+KFk3IbFP/plEHNrxdXnOP9324CpXC53dz4lw8z4ZLLexx76rfvEPzFGQwMbG1T9uXLqPyjOT2OjfchGC6TGbd+NnZeHx16n50P3ocVCmG6XL0es+s+Lvr2Hw9778FIRWuN9vvDy+ebm7F9vo76Qy2dl9TH1x2yo8vwz5/v+wCmGU6NFReH6w4Vl2IUhe93qjtUVBxOqRUVYcTVIjJyuqcQfnvqt2z4/1ez7bpHqa6oztBvRsgEXT+DLrnyqvQbuu1Q2Ng/iFhWO21t9Z0FU1uHf8m22+JGG+TljceTN4myMdd2XiHnmYzbPWpQ5hxoC3H83UaO7m/g+MFGQgGb3AJXpIRBOZMuK8V0Z+b3GBVT+ZLmy35ypxSj8k1cpR5GfWnGoIuFcTPC5R+meueyaPk3ehVSAJV/NIdTvEPTgXpGVU3qVUhB+LzKbp1Ne10zudNLOp3XhEsu58YfPEx97btMmjm712P2tQ/BeehgsEPgxIuhXosx+rCbm8NjfT4IBvvcv5GfHxNDRnER7gkTMC+9NE4gdQig8JgOUWQU5Kf9YtnUHo6uluQOXiRZSA+JfgalhLYgzdEbrTXB4NnOUaVIzaWOUgIdmGY+Hs9k8j1TGVO6uJNgysurxDCS9xmlg7YLQY69c4aj+09T/95ZrJCNpziHS68cz4zqciZcMgrTzLwQ9QfEMzWsUKZBzoTCIRELthVOb0ydU53QB0rlH83pU0TFkzuluNdziho1U9mHkF601tgXLsSiPd2KMcaiQc3haFGnytQt6NZ+qge73R3VqCMCKGdiZUcV6pLiSLQoPoIUEUWFhSi3e3B+EQnS3B5OW4/KHZxv8EJ6SfQzaMDY1oDSfLYdLlTZNQ0XjTZZVufWNrk5Y/F4JlM6+jOxqFJUNLndYxxTQqG1JUDdgdPU7W/g1B+asG1NYWkusxZXMn1uOeOmlwx6X1m/eKaEdGHb4TeTclBuXxg4diAQi/Z0iKL4KtQ9VKaOiiafr1/vkFFUFGnNEV5R5p4ymbzo6rKSeFFU1FkMFRej8vIc88GeDqKRqeIcEfvDkebmfZw79zqjR9dQUjI3+R1oK1zEtIf95efP6NG35PfX09b2EfFLcw0jN1acctSoheTHCaa8vImYZl4azjYz+M62RQTUaT460gQaSio8VP3pZKZXl1MxpWhIPxP8QQuXoXAPQhSsL0RMDQN05OJpGEP7ZhLCaNsO+4XivUM9teZo6ahW3fF8C7q95wUBUVRubqfl9WbZGHKmTevwDsVHiKKiKCKcjMLC8CozAYCmtiaKcopwZcCIK6QXrW0sqxXLukAodAHLOk/IuoAVuhDeZl3ACkW2WRfwtx7nTOMrhEWNwehRC3G7Ryd30NITgIJ3v0cweI5zTXvprX6J2z2GfM9kRpXMI2/cV8j3TCbPM5l8z2RycsozVyw1AzSfbuXovtMc3X+ahmMtAJROKGDB9VOZMbeC0gkFjvlS5Q9aQx6VAhFTw4KomOpafE4YGFprdFtbODXW0tzNUN1JAPXgHbLPn++7X5lhdEqTGcVF5JbPiEuRdTVUx4mh4mKM3MSK2Qn90xxolhRfhtBaY9ttceKnQ/B03I/cWufj7l8gFBkT/9quqbG+MM1oyxE7dnv+wmFyckqTOwl3IFyrrPUIgcDZuP0pSksXM7HyG+RFIk4uV+99Xp2O1pqzH1+gbv/pTm1cKqYUceVXpjOjuoJRY/suxDlUtAWtIfdLgYipYYEtYqobOhTqkiZr6by6rIuhuqt3qD8jtcrPD0eFIikw97hxmJdcHDZPFxV172ofly4z8vPl38ohNLeLmIrHtgNdxE+cyOkSDepZBEW2WRewrNZOVbP7wjDyMM0CXK4CTLMQl1lATk4ZpjkFl1mA6SqM3BaEx5mFmK6C8DazAJerEDNy3zQ9KGXQ3LyPfftvxraDGIYb75xNyaf6nvgTyCmEzz3XbX/Tp31vYKlDhxBt43J0X7gPXtOnkTYuMyJtXKrLKR7TexsXp+APSGRKSBM64pnK5jRf6/79tO59g/yFC8ivro4YqVtjHezjvUM9m6c7utrbzc3Y/RmpXa4OI3XUOzRhQmcB1MVQHb+6zGlGamFgNLU3UZqXZLTCQWhtYVmtsUhOp8hPL+mvsPg5H3e/QwjFF3XsC6XcceInLGpcrmJyc8fHRE9Y6MSJHldEBHV93szPSL2jkpK5zK3enppnyg7FDOhp2d8Qo23NJx+2cHR/A3X7T+NrDLdxqbxkFN5rJzKtKvE2Lk7BH7SGvCwCiJgaFkQjU0aWemFa9+/n+Iqbw5WplcIoKMD2+8OP+8AoLOzkHXJPnkRefJqsJ+9Q1Ejt8Tgm5y8MHc3tzUwvmT5oxwunvvw9e366pL+6RYa6CKFQ6AK27U/wyAammR+L4ETFjdtT2i3KExM9nbYVxkWBCjCM7LjglpTMTU302J1LI6S8vyHAtmw+OtzE0f2nqTtwmtbmAIYr3MZlwRemMm1OOXmF2fvl0B+0yZPIlJAOst0z1br3jQ7hpDU5M6ZTcOWiXkWRWRT2EYmRWkiVpvamPmtMhcVPIJa66pr+6j3y08s26wKJ9p0yzfxY6iqa/srNHUu+a3qcICrsLoLiIkbR5w1jeK3CHDS03Wk1X7ZghWxOfnCOo/sb+DCujcuUWWOYXl3O1Nll5KTYxsUp+AMhSfMJ6SHbxVT+wgXhO0qhcnMZu3Yt+dVSjVoYGLYdijMvd432dESDAsEWPldwjkuCe3jn3e/26gPSOpTQcQ0jJ5K26hAzOe7RmJ5JnaJB3dJf8Z6f2LZ8VBZexIcdthXur5gFBAMW9bXRNi5nCLRZ5OSZTJldxoy5kTYuDkiHpRt/0KKiaOgjayKmhgG2ld2eqfzqaoyCAnIuuoixa78vQmqEkeyS966prq7bbLvv0hIdmMwvsMgNnaS1NRATPLm55T2anHuM/JiFsfuGMfQf6EKasUOOjkwF/CGOH+zcxiWvwM2MuRVMz3AbF6cgBnQhbcRW82Vz2ss08cyeLUIqC4guee/Z7JxA+ss632m5u2X1s1gghuoU4Yned+dVdtvWb/rLLKCuuZ5lzy/j7xY/zJXTrsvo70zIUjLQTiZV2i4E+fDtM9Ttb+DE+2exQ5r84hwuu3I80+eWU3nxKIwhLmA5mLSJZ0pIF8OiaKfWMMhtCEYSzlryXo7HM3XAS97TRXMg3EpG+vIJvTLAdjLp5kJze0xAxbdxmX3NRGZUh9u4qBH6+Smr+YS0ke2eKQBsWwyycciS98wTFVNSZ0rolSE0oPvOtkWKaDbw8dFm0DBqbD5VSyczo7qc8slD28bFKfgDFh4RU0I6iHmmsqhdQTe0BrL3g0FrHfP99GR87iZyekl/hQVUckvee/Lv5HjGRKI9hYmJIDM/q5a8pwNpciz0ix0aVAN6U0NrWEDta6DhuA+AMZUFLPjCNGZUlzuqjYsTsG2NP2hJmk9ID3oYeKY0DOqHltOWvOfljkvO8yNL3lMm2uRYxJTQK3ZmPVNaa85+dCFcA2r/aRpPdbRxWfTVGUyvKndsGxcn0B4KX/vEgO4QEuksnnL38TTPJx57CDxTic4x4XOx7XAPrD72UVIyn8LCS/pd8t7N8+OoJe8FWdXwdDjT1N6E23DjcTm/ZYYwRGgr7Wk+rTWnT/hiAiq+jcsf3Xgx06vLKSrNS+sxhyv+YDgr43HAisURL6bOnHmFt9+5lWhUwTA8sfouEyeF+7P9+pV/6JR2iR8TRS+wwFSoV1P7R9Xa6vdY3V5j28xe1caRhm9y9Ezm31SJzjGZc7EfPs/HOT/l/Vf/V5/7SASlXF1WfRVGlrxXdNvWswjKlyXvI4BoXz6J7Am9kiYDurY1n9Q1xwSU72xcG5clk5jmLcu6Ni5OICamxDM19DQ2vkpHekZRVHQFxcVzAKitrQVg0iSL5uZ9kXGdx0S58PrHuEo95E5OLWXQ0vJOv8fqSnPDpxx5+/dcuuhqCksz32cs0Tkmcy5nt28n74rLyJ83r899jBnzWSrKl/ZZ98cwcuQCKfRLU1vf1c8FAXvgkSnbsjl1uIm6Lm1cJl9eyoIvTmOat4y8Avmilgr+QFhMiWfKARQVz4ZTAAaGkcPFF62NpaN+99snAbj2j72duoXHj4ny0c9+jyd/DKMvvjil+XTtTN7TsbpS53uD37z2IZ/94u2Mv+jSlI6fzjkmcy7vP/szxoxaTMU31vS5j2lTv5t1vbEEZ9JfKxlBCNeZSvxCbQVt6j84S93+0+E2LheCuHLi2rjMGj5tXJxAWyQylZ8z9L/ToZ/BEFNUeDkAY8d+kUkTb+7xQj2Y3cIHcqwOz9TgqPNE55jUuWjdo2dqOHRqF5xJc3szU0umDvU0BCeTQJovGLA4UdvI0X2nOf5uRxuXqXPKmFFdwaSZpcOyjYsT6PBMDf3vd8SLqSgVFZ/v80I9mN3Ckz2WtsNvqMGsM5XoHBM+F9vuVBnhk7pmTh06R+Uloxk3Pfs6tQvOp6m9SVbyDXM6f44MIArZxYAe3V/FlGL85wPU7TvN8dq4Ni7zKphRXcHES0cP+zYuTqA1EPVMDf3vWsTUMGBYFO3UOjb/T+qaee4f92OFbAxTMfPqCRSNkRVXQjrRTK6rYnTrRey/cGKoJyNkAF+jn9rdH2FbeuCfI77r4fAk0Cc67S9KfkkOly0az4zqciaMsDYuTkA8U0Ja6Wh0PPRvqAETV7Tz0N5PsCL1Q2xL8+4rp4ZwYsJwpYYvw3HY8/qRoZ6KkGEG/jnyLXgbeLv7e2Tm1RO45s8uHbFtXJxAm6T5hHSS7ZEprSPf9AyDxlPn+cPvPwHCFirDZfCF785h7LTiIZyhMNz46PzHLPvFV7mv5j6+NONLQz0dIQN8+mELu37yDrZlY5gD+ByxgrBhKiz+n/BHd3fan2kaXLZovAipIUZKIwhpZSiKdqaVyPxbAjm8+k/7ceeafHbFZbSc8Q/c6yAIfXDhQgtBs53RRSXk5MnH4HBk0uWlfGVN9cA9U8EQGG2QC+S5Ut+fkHaiaT6JTAlpIdsjU2hNa14Zr9VNBg/csKaa0eMKhnpWwjCmuU368o0Exk0vGbjosSMdEuLayaS0PyHtOCkylaVXXyEeewhW86UTX6Of/VV3Y2mDG/5ChJSQeaQvn9AvOvy5mu52MkL6aAtaGApyHGD8H/oZCCkTjUwZWdjo+Py5dn7+o3exzDyumXGKMZWFQz0lYQQQFVNStFPolciX1HS0kxEyQ2vAwuM2HdHxQsTUMCBbPVOtLQF+/k/78fuCeN/5MaWFwaGekjBCaG4Pp/mKc2Vhg9ALOvy5KpEp5+IPWo5I8YGIqWFBNnqm/OfDQur8uTa+8J3LKPEd61S0UxAySVN7E4XuQtzSxFrojZhnyhkXa6E7bQHLETWmQMTUsCBaZypbxFTbhSC/+OcDNJ/284XvzmH8tCIge+YvZD/Sl0/oF0nzOR5/0HLESj4QMTUsyCbPVMAf4vkfvc3Zjy9w3R2zmXhZaaRgJ0hoShgsmtubxXwu9I0Y0B2PP2iRL2k+IV3YWZLmC7SF+OW/vs2ZEz4+d+sspswcA8QV7XSAiVAYGYiYEvpFIlOOxy9pPiGd6CwwoIcCFi/82zt8UtfMn357JtOryjuejFVAFzElDArHNKMAACAASURBVA6S5hP6JSampByjU2kTA7qQTpzumbKCNr/a9C6nDjWxZNUVXDSvovOAaGRNIlPCICGRKaFfYmk+Z36uCh2lEZxAQu8SpdTnlVJ/UEodUUqt7eH5yUqpXyul9iul3lFKXZ/+qQq9oXVUjDjvj94K2fzvzQc58d5Z/njFZVxaM677oFiaz3nzF4YfITuEL+gTMSX0jaT5HE9WGdCVUibwY+A64Argz5RSV3QZdj/wM611NfAN4CfpnqjQO9q2UYbhuMiObdn8fz+t5dg7Z1j8jUu44qoJPY4Tz5QwmERrTEmaT+iTaGkEMaA7lragRZ5D0nyJJIMXAke01nUASqmngBuA9+LGaCBa/a4E+CidkxT6xrZtx/mlPjrSxG/+4xCNp85z1dcuYvZnJ/Y+WMSUMIhExZREpoQ+0eKZcjp+B6X5EnmXVAL1cY9PAjVdxqwDXlJK3QkUAH+SltkJCWFbFspBoehP6pr5+aP7sS2NYaj+G4OKAV0YRKSVjJAQES+npPmcidZ6WJZG+DNgq9Z6InA9sF31YOBRSn1HKfWmUurN06dPp+nQgrZtDAc0eoxy6tA5bCsskDSaU4fO9f0CMaALg4g0ORYSQupMOZqAZWNrsqo0wilgUtzjiZFt8Xwb+BmA1vo1IA8o67ojrfXjWuv5Wuv55eXlXZ8WBkjUM+UUKi8ZHfOSm6ZB5SWj+xyvxYAuDCLimRISItZORj6XnIg/EBa7TknzJfIueQO4WCk1TSmVQ9hg/osuY04ASwCUUpcTFlMSehokbNt2VJpv3PQSpswqw51rcsOa6sTTfBKZEgYB8UwJCSF1phyNPxgRU9mS5tNah4DvAS8C7xNetVerlFqvlPpyZNj/AG5TSr0N/AewSsfCDUKm0bblOAN6fpEbd57Zv5AC8UwJg0pTexMu5aLAXTDUUxGcjKT5HI3TIlMJSW6t9QvAC122/U3c/feAq9I7NSFRbMtZaT4IL+9MWBpFxJR4poTBIFr9XN5vQp9InSlHE41MZZNnSnA42rYxnPYHr0k8bRddNSMXN2EQkOrnQkLYEplyMm3ZluYTnI+2LWdGphLURrGEsBjQhUFA+vIJCaElMuVk/IHwl/DhVhpBGEKcWLSTZCxzWiJTwuDR1N4kkSmhfyTN52hiBnRJ8wnpwnZYaQQgOdNU1DMlBnRhEGhub2ZUnogpoR/EgO5oWgPh0hXimRLShrYtDNMZb6goYS0lninBWWitw2m+HEnzCf0QqzPlrM9WIYx4poS047SinUBSkSlpdCwMFv6Qn6AdFM+U0D+xdjJSZ8qJOK00gsOuwMJAcGKaT6MT10ZiQBcGCWklIyRMLM0nn0tOxB8Mi10RU0La0JbzinaSTMnWmAE9IzMRhBhS/VxIGDGgO5qOOlPOuPY5YxZCStgOrDOlB1AbwWnRNWH4EY1MSZpP6JeYZ0rSfE6kLWjhcZuOKb4rV69hgDM9UzrxxXxiQBcGCYlMCQkjq/kcjT9gOcZ8DiKmhgVOFFMakiiNELkVb4KQYWKeKSmNIPSHpPkcTWvAcoxfCkRMDQtsBzY6RifTay+6mi9jsxEEIC7NJ6URhP6IeTmdc8EWOmgLWo7xS4GIqWGBbdsop9WZSsaALmk+YZBobm+mwF2A23QP9VQEpyN1phyNPyhpPiHNaCe2k0mqNIIY0IXBQVrJCAkjaT5H45c0n5BunOiZSqbTsbalaKcwODS1N1GcUzzU0xCyATGgO5pwZMo5Ky0ddgUWBoIzi3YmoY2kArowSDS3N0tkSkgMiUw5mnBpBOdc95wzE2HAhIt2OuwPfkBFO0VMCZlFxJSQMDEx5Zzoh9CBPyhpPiHN2A70TGk9AM+UiCkhwzS1N0nBTiExpJ2Mo2mVOlNCunGkZwpIotNx+NaR5yAMFyzbwhfwSY0pITFsKyyk5EueI2kLWORJZEpIJ04UU+F2MgmOtWNVOzM1HUGgJdCCRkuaT0gMbYn53MFImk9IO44s2kkyX+iikSkRU0LmkL58QlLYIfFLOZSgZROyNfmS5hPSiROLdiZVtTNStFM8U0Imkb58QlLYtqzkcyj+YNjPJmk+Ia04sWhnuDSCeKYE5xDryydiSkgESfM5lrZAWEyJAV1IK070TCVTGkFr8UwJmUfSfEJS2CH5gudQWqNiSiJTQjqxLQvlsHC01kl4pqQCujAISJpPSArbEs+UQ4mm+URMCWlF2zaG6bR/yqSqdgKgxIAuZJCm9iZMZVLoLhzqqQjZgKT5HEvMMyVpPiGdOLFoZ1L9ZGypgC5knub2ZkpyS2Shg5AYYkB3LG2S5hMyQdgz5Zw3FSRXAV2LAV0YBKT6uZAUdkgiUw4lGpmS0ghCWgl7ppz1T5lMZQSkaKcwCEhfPiEptCWRKYcinikhI4RLIzjnTRUl6aKdoqWEDCKRKSEpbBFTTsUfkDpTQprRWqO180ojhFfzJVdnymnnIAwvmtqbJDIlJI4Y0B1LLDIlaT4hXWgdNm87zoCOTqI3nxjQhcwjaT4hKaQ0gmPxiwFdSDe2FWnF4jQxlVRlhGiaz2HnIAwb/CE/7Va7pPmExLEtWRTjUKSdjJB2tB1+UxkO6803kDSfRKaETCEFO4WkkTSfY/EHLXJcBqaDahOKmMpyoikyx0WmkkFL0U4hs0hfPiFpxIDuWNoClqPKIoCIqazHtp3pmUqmzpREpoRMI335hKSxQ+KZcij+oOUovxSImMp6HB2ZEgO64BCiaT4RU0LCaFvSfA7FH7RFTAnpxbYinimnhaM1JK6mIrdiQBcyhHimhKSRNJ9j8QdCjjKfg4iprMepkamwAT3RwdHIVMamI4xwxDMlJI225AueQ/EHLUfVmAIRU1mP7VAxlVRtBCnaKWSYpvYmPC4POWbOUE9FyBbEM+VY/IEs9UwppT6vlPqDUuqIUmptL2O+rpR6TylVq5T6f9M7TaE3tGMN6IlHpsQzJWQaKdgpJI2k+RyLP2g7LjLVr+xWSpnAj4E/BU4CbyilfqG1fi9uzMXAXwFXaa3PKaUqMjVhoTN2pM6UI6M6Caup6HgHnoMwLJBWMkLSSJ0px9KWpav5FgJHtNZ1WusA8BRwQ5cxtwE/1lqfA9BaN6R3mkJvxCJTTizamcxgEM+UkDGkybGQNLYtkSmHkq1pvkqgPu7xyci2eC4BLlFK/U4p9Xul1OfTNUGhb5xqQE9KTUUM6AlXTBeEJJE0n5A0dkjElENxogE9Xe46F3Ax8FlgIvAbpdRsrXVT/CCl1HeA7wBMnjw5TYce2Ti1aGeYJNvJOPIchOGARKaEpJE0n2PxB6ysLI1wCpgU93hiZFs8J4FfaK2DWusPgUOExVUntNaPa63na63nl5eXD3TOQhzROlNOi0wlZ0CXCuhC5rC1TUt7i4gpITnEgO5IQpZNwMrOop1vABcrpaYppXKAbwC/6DLmOcJRKZRSZYTTfnVpnKfQC871TCWT5ouKKWcJQmF44Av40GhJ8wnJIZEpR9IWCl/zPDnOul70m+bTWoeUUt8DXgRM4Kda61ql1HrgTa31LyLPLVVKvQdYwL1a68ZMTlwI41jPFEl4oKJ1piQwJWQAKdgpDAjbyso6U8FgkJMnT9LW1jbUU8kIlq3Z/OXxjMpv5f3338/IMfLy8pg4cSJutzvh1yT0TtFavwC80GXb38Td18A9kR9hEIl5phwW1dFJ1OzsqIAuakpIP9LkWBgQtpWVPs6TJ09SVFTE1KlTh+WinkDIwv7Ex8TR+ZQWpL8Ir9aaxsZGTp48ybRp0xJ+Xfa9U4RO6JhnymHh6GQ8U2JAFzKI9OUTBkSWpvna2toYM2bMsBRSAFGLrZGh01NKMWbMmKQje3L1ynJikSnTaf+USYSmxIAuZBBJ8wkDIosN6MNVSAHYkS/fRgbPcSC/P6ddgYUkcapnKryaL8nSCMP4A0AYOpraJM0nDIAs9Uw5lXXr1rFx48aU9xPtPtZTZOqxxx7jySef5MCBA9x+++09vl5rzV133cVFF13EnDlz2LdvX8pzAhFTWY+2HZrmAynaKTiCpvYmDGVQlFM01FMRsoksTfMNd/qKTO3evZvFixfz6quvsnjx4h5f/6tf/YrDhw9z+PBhHn/8cVavXp2WeYmYynKcWrQzucoI4pkSMkdzezMlOSWOW6QhOJwsNaAPhLeOn+PHvz7CW8fPpWV/27ZtY86cOXi9Xm6++eZuz2/evJkFCxbg9XpZvnw5ra2tAOzcuZNZs2bh9XpjYqi2tpaFCxdSVVXFnDlzOHL4EAAqLjT16KOPUlVVxbPPPsvy5ct54IEHePjhh7njjju6HfvnP/85f/7nf45SiiuvvJKmpiY+/vjjlM9ZYphZju3QOlNJqqnwrUSmhAwg1c+FAWGHsj4y9cPna3nvo5Y+x/jagnzwiQ9bh1Nnl40roiiv95IAV0wo5oEvzez1+draWh566CH27NlDWVkZZ8+e7TZm2bJl3HbbbQDcf//9bNmyhTvvvJP169fz4osvUllZSVNTOD2/adMm7r77br75zW8SCAQ40+LnTJvulOZbs2YNK1as4JZbbuH5559n4cKF7N27t8f5nTp1ikmTOuqQT5w4kVOnTjF+/Pg+f0/9MTJk9zCmI83nxH/KBMWRGNCFDNIcaBYxJSSPHhmeqZa2UOwj2Nbhx6nw8ssvc+ONN1JWVgZAaWlptzEHDx7k6quvZvbs2ezYsYPa2loArrrqKlatWsXmzZuxIivVFy1axCOPPMKGDRs4fvw4uXl5QPc03759+/B6vbS0tDBq1OAvNhn+75RhjrMN6EkMRjxTQmZobm9mbP7YoZ6GkE1oHfZyOtGLmgR9RZCivHX8HN984vcEQzZul8E/f6OaeVNGZ3Req1at4rnnnsPr9bJ161ZeeeUVIByFev3119m1axfz5s3jrbfe4qabbqKmpoZdu3Zx/fXXs+HRH3FR1ZWxyFRDQwNLly6loaGBvLw8nnrqKXw+H1VVVTz99NPMmDGj07ErKyupr6+PPT558iSVlZUpn5OzrsBC0jjVMwUkbUCXyJSQCSTNJyRN7DMpu8VUIsybMpodt17JPUsvZcetV6YspK699lp27txJY2O4CUpPaT6fz8f48eMJBoPs2LEjtv3o0aPU1NSwfv16ysvLqa+vp66ujunTp3PXXXdxww03UHvwXaDjy3dFRQUHDhxg7ty57N27lxUrVsRW9HUVUgBf/vKX2bZtG1prfv/731NSUpJyig8kMpX1aMc2OtZiQBccQXN7s9SYEpLDjqS6sjwylSjzpoxOWzRq5syZ3HfffVxzzTWYpkl1dTVbt27tNObBBx+kpqaG8vJyampq8Pl8ANx7770cPnwYrTVLlizB6/WyYcMGtm/fjtvtZty4cfzjpjVYSnVK81mWRWNjI2VlZezZs4d77um9Gcv111/PCy+8wEUXXUR+fj5PPvlkWs5bxFSW0xGZctgfvSaJEuiRW4lMCWmm3WrHH/KLmBKSI+JFHSliKt2sXLmSlStXdtq2bt262P3Vq1f3WJLgmWee6bZt7dq1rF27Nvb4oyY/51oDncaYpslrr70GwEsvvdTn3JRS/PjHP+73HJJFQgFZjmM9U8kMjlZhSziWJQiJIQU7hQGhI2JqBKT5sg1b64xWPx8ozroCC0ljO1RMoXUSgaaIAT1TzZaEEYu0khEGhESmHEu0hIPTcNgVWEiWaGkEpxnQtSbhQFM0uiZpPiHdSJNjYUDExJQ4YZyGbWtHrvx21hVYSBrnFu0ElbCaitw6TBAK2U80MiVpPiEpYmk++UxyGpLmEzKCcz1TSaT5JDIlZIjmgESmhAEgaT7HoiXNJ2QCp4opkkjzxTxTIqaENBNN80lkSkgKMaA7FolMCRnBtpzpmQKSKI0g7WSEzNDU1kSemUeeK2+opyJkE7E6U+KZShfr1q1j48aNKe/H7qO7xmOPPRYr2Hn77bf3OOaDDz5g0aJF5ObmpmU+URx4BRaSIbaaz2GeqaT6HEuaT8gQUv1cGBCS5nMsfUWmdu/ezeLFi3n11VdZvHhxj2NKS0v5l3/5F/7yL/8yrfMSMZXlaKe2k0lKTUVunXYOQtYj1c+FATGC2skAUL8Xdv9D+DYNbNu2jTlz5uD1ern55pu7Pb9582YWLFiA1+tl+fLltLa2ArBz505mzZqF1+uNiaHa2loWLlxIVVUVc+bM4djRIxhdTFOPPvooVVVVPPvssyxfvpwHHniAhx9+mDvuuKPbsSsqKliwYAFutzst5xpFYphZjlM9U0lZpqLnIJEpIc00tTeJmBKSJxaZctbnatL8ai188m7fY9pb4NODYQGpDBg7C3KLex8/bjZc97e9Pl1bW8tDDz3Enj17KCsr67E337Jly7jtttsAuP/++9myZQt33nkn69ev58UXX6SyspKmpvBK3E2bNnH33XfzzW9+k0AgwDsnznYzoK9Zs4YVK1Zwyy238Pzzz7Nw4UL27k2PMEyULH+nCHbkj145bQlvMu1k6CMJLggpIGk+YUCMJM9UW3NHJE7b4ccp8PLLL3PjjTdSVlYGhNNqXTl48CBXX301s2fPZseOHdTW1gJw1VVXsWrVKjZv3owV8QMvWrSIRx55hA0bNnDs2DFy8vJ6TPPt27cPr9dLS0sLo0YN/heoEfBOGd5o20YZhuOiOhoSL9qpRUwJmUHSfMKAGC6r+fqIIMWo3wv//mWwAmDmwPInYNLCjE5r1apVPPfcc3i9XrZu3corr7wChKNQr7/+Ort27WLevHm89dZb3HTTTdTU1LBr1y6+8IUv8D8f+gduuH5pbF8NDQ0sXbqUhoYG8vLyeOqpp/D5fFRVVfH0008zY8aMjJ5LFIeFM4RksW3beX4pCLeTSXRsX8szBGGA2NqmOdAskSkheUaSAX3SQlj5C7j2vvBtikLq2muvZefOnTQ2NgL0mObz+XyMHz+eYDDIjh07YtuPHj1KTU0N69evp7y8nPr6eurq6pg+fTp33XUXX/rSlzn8fm2nyFRFRQUHDhxg7ty57N27lxUrVsRW9A2WkAKJTGU94ciUQ//gkymN4ERBKGQ1voAPW9sSmRKSZ6QZ0CctTFs0aubMmdx3331cc801mKZJdXU1W7du7TTmwQcfpKamhvLycmpqavD5fADce++9HD58GK01S5Yswev1smHDBrZv347b7aZi7Fh+sHFTtzSfZVk0NjZSVlbGnj17uOeee3qd3yeffML8+fNpaWnBMAz+6Z/+iffee4/i4j58YgkgYirL0bblOPM5JLeYD20nPlYQEqSlvQWAUXkipoQkiXmmRoiYSjMrV65k5cqVnbatW7cudn/16tWsXr262+ueeeaZbtvWrl3L2rVrAWgLWhz61NfNgG6aJq+99hoAL730Up9zGzduHCdPnkzkNJLCeVdhISlsy8YwHfjPmMxyPolMCRkg2pdPIlNC0oykNF8WYUcKPEsFdCHt2A5N82kS90xp8UwJGSAqpopzUgvfCyOQ4WJAH2ZEm2VIbz4h7WjbcqgBHfFMCUOKRKaEASORKUcSjUw5bfU6iJjKeqKlERxJEmk+5/1pCNlOtMmxiCkhaWJiSmzFTkLSfELGsC1niqlkDeiS5hPSTVN7EwpFUU7RUE9FyDYkzedIbEnzCZkinOZz3h+8TkJNaUnzCRmgqb2J4txiTAf+fQgOZ7i0kxlm2BE11bU3nxOQd0qW49yinUnktcWALmQAqX4uDBgtab50s27dOjZu3JjSPqKRqd4uF4899lisYOftt9/e45gdO3YwZ84cZs+ezWc+8xnefvvtlOYUxYFXYSEZnOqZSqadDFo70lAoZDfSl08YMNE6U5LmcxS6H8/U7t27Wbx4Ma+++iqLFy/uccy0adN49dVXeffdd/nBD37Ad77znbTMzXlXYSEpbIcW7UzKVC69+YQMIJEpYcCMsNV8BxoO8MS7T3Cg4UBa9rdt2zbmzJmD1+vl5ptv7vb85s2bWbBgAV6vl+XLl9Pa2grAzp07mTVrFl6vNyaGamtrWbhwIVVVVXx20QJOfFjX7dry6KOPUlVVxbPPPsvy5ct54IEHePjhh7njjju6Hfszn/kMo0ePBuDKK69MWwFPiWFmOdq2MUwH/sEnURpBiwFdyADN7c1cPPrioZ6GkI0Mk3YyG/Zu4IOzH/Q55nzgPH8494dIbUDFpaMvpTCnsNfxl5VexvcXfr/X52tra3nooYfYs2cPZWVlPfbmW7ZsGbfddhsA999/P1u2bOHOO+9k/fr1vPjii1RWVtLUFC5tsmnTJu6++26++c1vcqyhmcYWf7dMxpo1a1ixYgW33HILzz//PAsXLmTv3r19njfAli1buO666/odlwgiprIc28FpvqQiUw48ByG7kTSfMGBGUDsZX9CHDhsz0Gh8QV+fYqo/Xn75ZW688UbKysoAKC0t7Tbm4MGD3H///TQ1NXH+/Hk+97nPAXDVVVexatUqvv71r7Ns2TIAFi1axMMPP8zJkyf5zJLrGDNhao/H3bdvH16vl5aWFkaN6j8i/etf/5otW7bw29/+doBn2hkRU1mOdrABPWE1JQZ0Ic0ErSCtoVZKckRMCQNgmKT5+oogRTnQcIDbXrqNoB3Ebbj526v/lqqKqozOa9WqVTz33HN4vV62bt3KK6+8AoSjUK+//jq7du1i3rx5vPXWW9x0003U1NSwa9cu/vz/WsYDGx7liq99KbavhoYGli5dSkNDA3l5eTz11FP4fD6qqqp4+umnmTFjRrfjv/POO9x666386le/YsyYMWk5JwdehYVksC1neqY0OnF9JAZ0Ic1I9XMhJUZQnamqiio2L93M96q/x+alm1MWUtdeey07d+6ksbERoMc0n8/nY/z48QSDQXbs2BHbfvToUWpqali/fj3l5eXU19dTV1fH9OnTueuuu/jTz3+BQ+/XdtpXRUUFBw4cYO7cuezdu5cVK1bEVvT1JKROnDjBsmXL2L59O5dccklK5xqPRKaynHBkyoF/8Eks5xPPlJBuomKqJE8iU8IAGCaRqUSpqqhKWzRq5syZ3HfffVxzzTWYpkl1dTVbt27tNObBBx+kpqaG8vJyampq8Pl8ANx7770cPnwYrTVLlizB6/WyYcMGtm/fjtvtpqS0nO/c9T+6HdOyLBobGykrK2PPnj3cc889vc5v/fr1NDY28t3vfhcAl8vFm2++mfJ5i5jKcpxaGiGpNJ94poQ0I5EpISWknUxKrFy5kpUrV3batm7dutj91atXs3r16m6ve+aZZ7ptW7t2LWvXrgXg6OnzPR7PNE1ee+01AF566aU+5/bEE0/wxBNP9DlmICR0BVNKfV4p9Qel1BGl1No+xi1XSmml1Pz0TVHoC6cW7UzOgI5EpoS0In35hJSIpfmc99k6krG1dmRfPkhATCmlTODHwHXAFcCfKaWu6GFcEXA38Hq6Jyn0jpPrTCVuQLeTUF6C0D8SmRJSYoSl+bIF23ZmXz5ILDK1EDiita7TWgeAp4Abehj3ILABaEvj/IR+cGqdqXCwKeHmfCj5BiikkZhnSkojCANhBBnQswmdzZEpoBKoj3t8MrIthlJqLjBJa70rjXMTEsDRnqlEh4oBXUgzLe0t5Jq5eFyeoZ6KkI3E6kyJZ8pJ2Dq7I1N9osIhhX8Eulvsu4/9jlLqTaXUm6dPn0710ALOLdoJJGFARwzoQlqRgp1CStiRCuiS5nMUWe2ZAk4Bk+IeT4xsi1IEzAJeUUodA64EftGTCV1r/bjWer7Wen55efnAZy3EcGLRzmgzyuR682VsOsIIpKm9SfxSwsCRNJ/j0FpjO7gmYSJX4TeAi5VS05RSOcA3gF9En9RaN2uty7TWU7XWU4HfA1/WWqdeuEHoF0cW7Yym+BJ909u2eKaEtNLc3iyRKWHgxAzo8rmULtatW8fGjRsH/PrId/Q+/0kee+yxWMHO22+/vccxP//5z5kzZw5VVVXMnz8/be1k+n2naK1DwPeAF4H3gZ9prWuVUuuVUl9OyyyEAePEop3JaimQdjJCepHIlJASdkj8Ug7DjqipvtJ8u3fvZvHixbz66qssXry4xzFLlizh7bff5sCBA/z0pz/l1ltvTcv8Enq3aK1fAF7osu1vehn72dSnJSSKEw3osTRfoov5pDefkGbEMyWkhLZGVIqvdf9+Wve+Qf7CBeRXV6e8v23btrFx40aUUsyZM4ft27d3en7z5s08/vjjBAIBLrroIrZv305+fj47d+7khz/8IaZpUlJSwm9+8xtqa2v51re+RXt7AH8gyI6nfkaZd2an/T366KP8+7//O4cPH+bgwYMcO3aMCRMmsHv3bjZt2tRpbGFhRxPnCxcupC1tKNI7y3Fk0c7YSr7ESyM4domGkHVorWlpb5HIlDBwbGtYmM8/eeQR2t//oM8x1vnztH/wQcS7qsi97DLMOMHRldzLL2PcX/91r8/X1tby0EMPsWfPHsrKynrszbds2TJuu+02AO6//362bNnCnXfeyfr163nxxReprKykqSlc3mTTpk3cfffdLP/6N6itb2RCSW63/a1Zs4YVK1Zwyy238Pzzz7Nw4UL27t3b6xyfffZZ/uqv/oqGhgZ27UpPEQKHXYWFZAkX7XTYH30sz5fgeNt2rKlQyD7OB88T0iERU8LA0faIiUzZLS0dhiStw49T4OWXX+bGG2+krKwMgNLS0m5jDh48yNVXX83s2bPZsWMHtbXh5sVXXXUVq1atYvPmzVhW2Le2aNEiHnnkETb+/d/x0al68vPzezzuvn378Hq9tLS0MGpU33/7X/3qV/nggw947rnn+MEPfpDK6caQyFSWEy7a6SxNrEkuzRf2TDnrHITsRQp2Ciljh4ZFZKqvCFKU1v37OfGtW9DBIMrtZsLGv09Lqq8vVq1axXPPPYfX62Xr1q288sorQDgK9frrr7Nr1y7mnbnI5QAAIABJREFUzZvHW2+9xU033URNTQ3P/PwXfO/Pv86P/+3f+OLnl8b21dDQwNKlS2loaCAvL4+nnnoKn89HVVUVTz/9NDNmzOh1HosXL6auro4zZ87ExN9AkStYluNEz1QyBTsh4rGSyJSQJqQvn5AywyTNlwj51dVMfvKnlN91F5Of/GnKQuraa69l586dNDY2AvSY5vP5fIwfP55gMMiOHTti248ePUpNTQ3r16+nvLyc+vp66urqmD59Ond893t8dun1vHfw3U77qqio4MCBA8ydO5e9e/eyYsWK2Iq+noTUkSNHYr7effv20d7ezpgxY1I6Z5DIVNYTLtrprD/6jtV8iZZGEDElpA8RU0LKjDADen51ddqiUTNnzuS+++7jmmuuwTRNqqur2bp1a6cxDz74IDU1NZSXl1NTU4PP5wPg3nvv5fDhw2itWbJkCV6vlw0bNrB9+3YM00VxaTmP/LB7Ws6yLBobGykrK2PPnj3cc889vc7v6aefZtu2bbjdbjweD//5n/+ZFpuJiKksR1uWgw3oiY4XA7qQPiTNJ6TMCIpMZYKVK1eycuXKTtvWrVsXu7969WpWr17d7XXPPPNMt21r165l7dq1NLUGOHG2lTFjirqNMU2T1157DYCXXnqpz7l9//vf5/vf/34ip5EUDrsKC8liO9EzFTUzJmNAlxLoQpqIiimJTAkDxrakzpTDSKTO1FDirKuwkDTagWm+KIkKJI2WSsNC2oim+Ypyun+DFYSE0JYsinEYdrQCujO1lIipbEfbTm4nk+h48UwJ6aOpvYminCJcElkQBoqk+RyHRKaEjGJbzivamXQ7GVs8U0L6kFYyQspoSfM5DdsO3zpUS4mYyma01mjtwDSfTro2gnimhLTR3N4sYkpIDTs0olbzZQNaawylHFvgWcRUFqN1WKo7LjIV9Z8n3pzPuV83hKxD+vIJKWPbkuZzGHZETDkVZ12FhaTQkbin4zxTSaK1GNCF9CGRKSFlxICedtatW8fGjRsH/PpE3CCPPfZYrGDn7bff3ufYN954A5fLxX/9138NeE7xyLsli7EjvYscJ6aS7s0nBnQhfYiYElJGSiM4DlvrfrMdu3fvZvHixbz66qssXry413GWZfH973+fpUuX9jomWRx2FRaSIRqZMkxnhaOT7s2XwB+JICRC0A5yPnhe0nxCagyT3nyJ8kldM2/972N8Uteclv1t27aNOXPm4PV6ufnmm7s9v3nzZhYsWIDX62X58uW0trYCsHPnTmbNmoXX642JodraWhYuXMh11yzihmsXcfjw4W77e/TRR6mqquLZZ59l+fLlPPDAAzz88MPccccdPc7vRz/6EcuXL6eioiIt5wtSAT2rsW1neqY6KqAn6pmSyJSQHqSVjJAWhkk7md0/O8SZ+vN9jgn4Q5w5dT78ua2grLKQHE/v0qBsUiFXf/2SXp+vra3loYceYs+ePZSVlfXYm2/ZsmXcdtttANx///1s2bKFO++8k/Xr1/Piiy9SWVlJU1O4+O6mTZu4++67uXLpDQQCASaWerrtb82aNaxYsYJbbrmF559/noULF7J3794e53fq1CmeffZZfv3rX/PGG2/09atJCoddhYVkcKpnqsOAnuh4MaAL6UHElJAWRpABvd0f6vgCrCOPU+Dll1/mxhtvpKysDIDS0tJuYw4ePMjVV1/N7Nmz2bFjB7W1tQBcddVVrFq1is2bN2NFbCyLFi3ikUceYdO//CMfn6zH4+kupiDctNjr9dLS0sKoUb3//f/FX/wFGzZsSHsQQiJTWYxzxVSyaT7EgC6khWgrmeLc4iGeiZDV2CFw5Qz1LFKmrwhSlE/qmvn5o/uxLBvTNFj67ZmMm57ZNPmqVat47rnn8Hq9bN26lVdeeQUIR6Fef/11du3axbx583jrrbe46aabqKmp4cmnnubbf7acLU88zrXXXhvbV0NDA0uXLqWhoYG8vDyeeuopfD4fVVVVPP3008yYMaPTsd98802+8Y1vAHDmzBleeOEFXC4XX/nKV1I6J7mCZTFRA7rj0nxRElVTtp24WV0Q+kD68glpYZik+RJh3PQSblhTTc2Xp3PDmuqUhdS1117Lzp07aWxsBOgxzefz+Rg/fjzBYJAdO3bEth89epSamhrWr19PeXk59fX11NXVMX36dFZ8+w6WXv8F3nnnnU77qqio4MCBA8ydO5e9e/eyYsWK2Iq+rkIK4MMPP+TYsWMcO3aMr33ta/zkJz9JWUiBRKayGjsWmXLYH32SNTvDBnSHCkIhq5A0n5AWRlg7mXHTS9IWjZo5cyb33Xcf11xzDaZpUl1dzdatWzuNefDBB6mpqaG8vJyamhp8Ph8A9957L4cPH0ZrzZIlS/B6vWzYsIHt27djYTJu3Fj+n/UPdDumZVk0NjZSVlbGnj17uOeee9JyLskgYiqL0Q41oCfrmRIDupAuJDIlpIURFJnKBCtXrmTlypWdtq1bty52f/Xq1axevbrb65555plu29auXcvatWupPdXM6IIcSkd190yZpslrr70GwEsvvZTwPLuKvFRw1lVYSAqneqaSDU1J0U4hXTS1N+E23HhcPZtUBSEhpM6Uo9BaR4p2OvdLt1zBshjbdqZnKvl2Mlo8U0JaiBbslLplQkrYlnzBcxCacP3C/iqgDyXybsliYpEphxXtjJFwBXQpjSCkh+b2ZinYKaSOpPkchR1bIe7c64SIqSzGqUU7o6URkniBGNCFtNDU3iR+KSF1RpgB3elELykSmRIygmM9U0kX7RQDupAepC+fkBbskHimHIRthy8qhoPVlMOuwkIydDQ6duo3qCQ8U04ThEJW0tTeJGk+IXW0LWk+B2FLZErIJB2Njp31zziw0ggZm44wQtBaS5pPSA9iQE8769atY+PGjQN6bdQz1d9qvsceeyxWsPP222/vccwrr7xCSUkJVVVVVFVVsX79+gHNqSsSx8xiYp4px/mNomoqweG27WhjoZAdtIZaCdkhiUwJqSMGdEehExRTu3fv5oc//CG//OUvWbx4ca/jrr76an75y1+mdY5OuwoLSaBtZ6b5dJJaCjQ4ThAK2YYU7BTSxgjzTH106H1ef/ZnfHTo/bTsb9u2bcyZMwev18vNN9/c7fnNmzezYMECvF4vy5cvp7W1FYCdO3cya9YsvF5vTAzV1tZyzR99hq9/7mpq5ldz+PDhbvt79NFHqaqq4tlnn2X58uU88MADPPzww9xxxx1pOZ9EGDnvlmGIbTnbgJ5onk/bYkAXUicqpiQyJaTMMFnN9+utj9NwvK7PMe2trZw5/iFaa5RSlE2ZRm5+fq/jK6ZM549XfafX52tra3nooYfYs2cPZWVlPfbmW7ZsGbfddhsA999/P1u2bOHOO+9k/fr1vPjii1RWVtLUFP573rRpE7d/93tcufQGpo7Oxd3D5W7NmjWsWLGCW265heeff56FCxeyd+/eXuf42muv4fV6mTBhAhs3bmTmzJm9jk0Uh12FhWRwqmcqabR2trNQyAqa26Qvn5AmRpABvb31QiyNprWmvfVCSvt7+eWXufHGG/k/7Z17fBTnee+/7+7qgiQQRuImIa4Cm6skIJIJB0hFgh3q2kUpJI5ppOAQlwTiAzUtx+AcLnYa17S0jV0jCI6AkpBgCMeYYDgfE2JOwWBDMZWMQcgGhG0QrJC0Qki7O/OeP2avWl1WrC670vv9fPTZ2ZlnZp5h2J3fPs/zPm9ycjIA/fr1C7ApLi5mxowZTJw4kV27dlFSUgLA9OnTKSgoYOvWrWiuAVbTpk3jn195mTf+/V8ov3aVXr2ant3g7NmzZGRkUFNTQ9++zX8HTJ48matXr/LRRx+xbNmydpnkGFRkKqIJ3zSfu8Fa0DuomilFyFTblZhStBPdJDLVUgTJzReXLrBnw2o0pxOzxcKfL3uOlDFjO9SvgoIC9u/fT0ZGBkVFRRw7dgwwolCnTp3i4MGDTJkyhTNnzvDd736X0RMyeXP/Af7y8b+gsLCQ3Nxcz7EqKiqYM2cOFRUVxMbGsnv3bmw2G5mZmezdu5dRo0b5nbtPnz6e5blz5/KjH/2I27dve8Tf/RLhIY2ejS7Ds2mnm6AFkq6rmilFyKg0n6Ld0J3dQkwFQ8qYscx/4SWmL1jI/BdeCllI5ebmsmfPHqxWK0CTaT6bzcbgwYNxOBzs2rXLs76srIycnBzWr19P//79KS8v59NPP2Xo8BE8tegZ/uLxxzl//rzfsQYMGMC5c+eYPHkyp0+fZuHChZ4RfY2FFMCNGzc8P/hPnz6NruskJSWFdM2gIlMRTbg27XRpvODbTKFqphSho8SUot3oYaP5UsaMbbdo1Pjx41m9ejWzZs3CbDaTlZVFUVGRn82GDRvIycmhf//+5OTkYLPZAFi5ciWlpaVIKZk9ezYZGRm8/PLL/Gr7DoTJzPC0VNasXh1wTk3TsFqtJCcnc+LECVasWNGsf2+++Savv/46FouFXr16sXv37nbJjCgxFcG4m3aawmxuPkkbp5NRBeiKdqC6oZreUb2x9KBRWIoOopuk+bqK/Px88vPz/datXbvWs7xkyRKWLFkSsN++ffsC1q1atYr8Jc9irbUzIbXpH0pms5mTJ08CcOTIkRZ9W7p0KUuXLm3tEtpMeIU0FG0iXCNT3ulkgu+ALlQBuiJEVPdzRbug64DsUa0Rwh09AsYohdlTWNEWwnWi4zaj66gW6IpQUd3PFe2CNCL+PSnNF+7oevgPUorwp3DPJlwjU22dTkai5uZThE51fbWKTClCxzVKWn0nhQ9Syla7n3c16n9LBOOd6DjcbmMbW6BLVM2UImRUmk/RLqjIVNjRbdJ8QohHhRAXhRCXhRCrmti+QgjxsRDivBDiXSHEsPZ3VdEYT9POMCuU9E4n05bWCB3nj6JnUN1QrdJ8itDxRKZUzVS4oHeHyJQQwgy8BnwTGAc8KYQY18jsv4CpUspJwJvAP7a3o4pA9DBN83mnkwnWXobfNSgiCqfuxOawKTGlCB3dabx+9h6UNz8liaLziIQB38E8wbKBy1LKT6WUdmA38ISvgZTyj1LKOtfb94Eh7eumoilkNylAl1IVoCtCo8ZeA6geU4r7xHEPbpbAx/8Hjqwx1pUege2PK0HVTqxdu5aNGzfe177BRqYKCws9DTufeeaZZu2OHTtGZmYm48ePZ9asWfflU2OCiWOmAuU+768DOS3YPw0cCsUpRXDoenjWTLV9OhlUsaciJNwNO1VkStEsugbV5WC9DNYyuF3qXa4uh4D+eBI0O1w5DmnZXeGxwoWUElMQRVPHjx9n3bp1vP3228ycObNJm6qqKn70ox/xzjvvMHToUCoqKtrFx3ZNCgshFgJTgSalnhDih8APAYYOHdqep+6ReCc6Dq+aKQ9t6DOlAlOKUKhuUPPyKTC+S+qshkjyiCWXYKr8FLQGr210b0hOh6E5kLQQkkZBUjrcq4LffMcQUuZoGD6j666nk2i4WkPDp9XEjEwkZlif1ndohR07drBx40aEEEyaNImdO3f6bd+6dStbtmzBbreTnp7Ozp07iYuLY8+ePaxbtw6z2UxiYiLvvfceJSUlPLnwe2hOB2YBe/fuZfTo0X7H27RpE9u3b6e0tJTi4mKuXLlCSkoKx48fZ/PmzX62v/71r8nLy/NokAEDBoR8vRCcmPocSPN5P8S1zg8hxNeB1cAsKWVD4+0AUsotwBaAqVOntrFNtqIxYd8aIdgddD3srkERWVTVu6aSiVVpvh6B/a4hjqyX4fZlH9FUCvXVXjtTFPQbYYik0d8wXpPSIXk0xPdv/gdf/ltGRGr4jIiOSlUdKMP+xd0WbfR6J84bd12jqsEyKB5TbPPSIDolnr5/ETjnnZuSkhJefPFFTpw4QXJycpNz8+Xl5bF48WIA1qxZw7Zt21i2bBnr16/n8OHDpKamUlVlfKY3b97MU0//DX+9cCHJcWY01yh2X5YvX87ChQtZtGgRBw4cIDs7m9Onm07PXrp0CYfDwde+9jVsNhvPPvss3/ve91r6JwqKYMTUB8BoIcQIDBH1HeC7vgZCiCygEHhUStk+MTNFq4Rt0877KEBXoSlFKKg0XzdEc0L1tUYpOddfTaPf831SjcjShL/yCqakUdB3GJjvIwGTlh3RIqotyHqn9ztbut63IKZa4+jRo8yfP5/k5GQA+vXrF2BTXFzMmjVrqKqqora2lkceeQSA6dOnU1BQwIIFC8jLywNg2rRp/HTdBmqsNyn47rcDolJuzp49S0ZGBjU1NfTt2/z3gNPp5MyZM7z77rvcu3ePadOm8fDDDzNmzJj7vmYIQkxJKZ1CiKXAYcAMvCGlLBFCrAc+lFK+BbwCJAB7XF1Kr0kpHw/JM0WrhGufqbbOzSdlBAzVUIQ1Ks0XoUgJd281SsuVGRGmys9Ad3htYxKNtNzwGV6x5H6Nju+6awhjWooguWm4WsPtX/430qkjLCb6feehdkn1tURBQQH79+8nIyODoqIijh07BhhRqFOnTnHw4EGmTJnCmTNn+M6TT5I4dCznThxl7ty5FBYWkpub6zlWRUUFc+bMoaKigtjYWHbv3o3NZiMzM5O9e/cyapT/v8GQIUNISkoiPj6e+Ph4Zs6cyUcffdTxYgpASvkH4A+N1v3UZ/nrIXmhuC/CtmbqPubmUwXoilCoaqjCIizEWeK62hVFUzTUQqU7wlTmTclZy6ChxmtnjoZ+IyF5DDw41z8tF5ekfnR1ADHD+pD8g4ntVjOVm5vLvHnzWLFiBUlJSVRWVgZEp2w2G4MHD8bhcLBr1y5SU1MBKCsrIycnh5ycHA4dOkR5eTmVd+4wZNhwsjOWcqfiS86fP+8npgYMGMC5c+d47LHHKCoq4tVXXyU7O5u5c+c26d8TTzzB0qVLcTqd2O12Tp06xfLly0O6ZmjnAnRF5xKuNVMegv3eU007FSHi7n4e7vN3dWs0B1RdC0zJWS+D7Ut/28Q0QyRN+raPYEo31odZE+KeQMywPu0WjRo/fjyrV69m1qxZmM1msrKyKCoq8rPZsGEDOTk59O/fn5ycHGw2GwArV66ktLQUKSWzZ88mIyODn/3Dz3lj+w7iYqJJTRnM888/H3BOTdOwWq0kJydz4sQJVqxY0ax/Y8eO5dFHH2XSpEmYTCZ+8IMfMGHChJCvW4mpCMbTtFOEl5hq69x8SBl216CILFT3805CSqi9GThSzloKd654G14C9HoAkkbDyD/zpuSSRxuRp6heXXYJio4nPz+f/Px8v3Vr1671LC9ZsoQlS5YE7Ldv376Adc/93d/xF99bQtoDcTwQH93k+cxmMydPngTgyJEjrfq3cuVKVq5c2apdW1BiKoKRuoYQpvD7Ne7uMxVsuEnVTClCRM3L187U1/gIpcv+aTl7rdfOEgv9RsGAcTD2ca9gSkqHuMDCY4Wirbh/nIf73HxKTEUwuq5jModfREe2cTSfKkBXhEq1vZq0hLTWDRVenHYjmtQ4JWe9bESfPAjom2ZEmdIe9qbkktKhzxBV76joUHRPE+jwfkYoMRXBSF1HhGV9QRsbTckImBJcEdZU11czMXliV7sRfkgJNV8EpuSsl+HOVZA+PXvikgzBlP4NIy3njjA9MAKiYrvuGhQ9Gl03nifhPtGxElMRjNS1sCw+v6+mnWH+QVGEL1JKlea7V9VESs4lnhx1XjtLL0MgDZoE4/O8gqnfSJWWU4QlujvNF36POj+UmIpgdF0Pv4ad4NO0sy01U2F4HYqI4J7zHnbd3v0L0J0NRu+lxjVM1stGryY3wmQ0q0xKh2H/w5uSS0qH3inh/1RSKHxwp/lUZErRYehaeE7D0uYG6KBqphT3Tbdq2KnrRnfvpuqYqq6B1L228QMMgTTmUW+EKSkdHhgOlpguuwSFoj3RVQG6oqORuhZ+DTuh7Xk+XQ//T4oibHFPJRNRab66ykbNK931TGXgvOe1i4o36pdSJvv0ZBpljKDr1Q3Eo6LHsHbtWhISEnjuuefatJ9sQwF6YWEh0dHRZGVl8frrr1NYWBhg88orr7Br1y7AmFrmwoUL3Lp1q8lpb9qCElMRjAzTCYLbmuVTrREUoeARU9FhJqYc9d7JeH1TctbLUGf12gmzEU1KSoeRX/OZJiUdeg9Wnw1Fj8YbmWr9c3D8+HHWrVvH22+/zcyZM5u08e0xdeDAATZt2hSykAIlpiIaPUzFlHdqviAfAqoAXRECXZrm0zWovt5MWq4cfOepTBhkCKSHHgtMy5mjOt93hcJFeXk5V65cYfjw4aSlhd5iZMeOHWzcuBEhBJMmTWLnzp1+27du3cqWLVuw2+2kp6ezc+dO4uLi2LNnD+vWrcNsNpOYmMh7773Hxx+X8ONnFmNBQ9d19u7dGzDZ8aZNm9i+fTulpaUUFxdz5coVUlJSOH78OJs3b27Wz9/85jc8+eSTIV8vKDEV0UhNC/MC9CBMPSnBMLwORUTgjkz1je0gMSWlKy3XxEg5axloDV7b6N5GZGlINmQ+5T8hb0zvjvFPoWiGQ4cOcePGjRZtGhoauHnzJlJKhBAMHDiQmJjma+4GDRrEN7/5zWa3l5SU8OKLL3LixAmSk5OprKwMsMnLy2Px4sUArFmzhm3btrFs2TLWr1/P4cOHSU1NparK+Fxv37aVhU//DauWLcZut6NpWsDxli9fzsKFC1m0aBEHDhwgOzub06dPt3jddXV1vPPOO7z66qst2gWLElMRjDGaL/xqpiTuDujBGLd17hmFwh93ZCrkNJ+9rum03O1SqK/y2pksRu+lpHRIn+2NMCWNhoQB6v+yIqKor6/3/KiVUlJfX9+imGqNo0ePMn/+fJKTkwGaTKEVFxezZs0aqqqqqK2t5ZFHHgFg+vTpFBQUsGDBAvLy8gCY8pUcNm18GVFXSV5eXkBUys3Zs2fJyMigpqaGvn1b/2F14MABpk+f3i4pPlBiKqIJ15qpNhVNRcpcAYqwpaqhivioeKKCSZXpmjEqzrd5pfUy3L4MNdf9bXunGFGlCXk+gindaDtgVl+divCnpQiSm/LycrZv346maZjNZr71rW+1S6qvJQoKCti/fz8ZGRkUFRVx7NgxADZv3sypU6c4ePAgU6ZM4cyZMzzxrQWkT8jk4gfvMXfuXAoLC8nNzfUcq6Kigjlz5lBRUUFsbCy7d+/GZrORmZnJ3r17GTVqVJM+7N69u91SfKDEVEQTrmKqTQXoETJVgCJ8CZjkWEq4ezswJXe7FO58BprdaxvTxxBIw6f7pORck/HGJHT+xSgUnUxaWhr5+fntVjOVm5vLvHnzWLFiBUlJSVRWVgZEf2w2G4MHD8bhcLBr1y5SU1MBKCsrIycnh5ycHA4dOkR5eTm1WhTDho/kkWmZXLt2jfPnz/uJqQEDBnDu3Dkee+wxioqKePXVV8nOzmbu3LnN+lhdXc2f/vQn/uM//iOka/VFiakIRtfDvGYqGHRX3xxVM6VoK/a7YC2j6vYnJNrvwb4feqNMrtQfAKYoQxwlj4YHH/VPy8Unq7ScoseTlpbWbtGo8ePHs3r1ambNmoXZbCYrK4uioiI/mw0bNpCTk0P//v3JycnBZrMBxki70tJSpJTMnj2bjIwMVv10A/t+9xvie8UwaNAgnn/++YBzapqG1WolOTmZEydOsGLFihZ9/P3vf8+cOXOIj49vl2sGJaYiGl3XEWHYZ0q2oc+Ud+CfeqApmkBzQtXVwJFyty+D7QsAqgcPpK+uw73bRrfvSfO9YilpFCSmqbScQtGJ5Ofnk5+f77du7dq1nuUlS5awZMmSgP327dsXsO6Zn6xg8bIVpA9oPlJsNps5efIkAEeOHGnVv4KCAgoKClq1awvqGyaCkeE6nYwLEYya8kSmlJjqsUgJtRWBNUzWy0ZaTnd6bWP7GhGmkbM8Kbmqj18jtf8k+No/dd01KBSKDkHKyCipVWIqggnXmina0u3AXTMVCZ8WRWg02HxaClz2jpSzloHd5rUzxxhCacBDMPYxnyhTujEZbyPhXXX+Ffr2Surki1EoFJ2BLiVR4fica4QSUxGMrmlhKabaUjKlWiN0MzQH3LkSmJKzXoZa3343AvqmGQIpLdt/tFziEAiy5Yema9jstsiaSkahUASNLsN/kmNQYiqikWHaZ6otI/Skrpp2RhxSgu1LH8HkGilnvWwIKenTVC8uyacf0yhvhKnfCIjqFbIrNfYaJLJ7THKsUCgC0KVUaT5Fx6KHac2UbFtoyniJgF8ePY76av+03G2fNgOOu147Sy9DKA2aAOP/0iuYkkYZabkOJCInOVYoFEGjSxkRZSBKTEUwUtcRljC8hW2YTsZbgN5RzihaxGk3irwb1zBZL8PdCq+dMEHfoYZIGvZV/7Rcn1ToIlHfpfPyKRSKDkcVoCs6HF0P15opd5ovGGN3AXr4XUe3QdeNNgKNa5isl422A1L32sb3NwTSmDk+ESZXWs5y/1NMdBRKTCkUkcHatWtJSEjgueeeC3ofKaUrzRecmiosLCQ6OpqsrCxef/11CgsLA2yqq6tZuHAh165dw+l08txzz/H9738/aJ+aQ4mpCEbqOqYw7DPl0zwqCNu2hLEULXLvTtMj5SrLwFHntYuKM1JwKZkwcb5PlGkU9IosUaLSfApF90Vv42xjx48fZ926dbz99tvMnDmzSZvXXnuNcePGceDAAW7dusWDDz7IU089RXR0dEi+KjEVwehh2hqhLQP0PA0+w/A6whJHvTct55uSs16GutteO2GGB4YZImnETFfxtzstl9JtatTcYkpFphSK+6e6+ix37pzigQdySEycHPLxduzYwcaNGxFCMGnSJHbu3Om3fevWrWzZsgW73U56ejo7d+4kLi6OPXv2sG7dOsxmM4mJibz7x2NcvniBgr//CbrTga7r7N27N2Cy402bNrF9+3ZKS0spLi7mypULAqmvAAAbCklEQVQrpKSkcPz4cTZv3uxnK4TAZrMhpaS2tpZ+/fphaYdyGSWmIphwb9oZFG3olt5j0HVj0t3GI+Wsl41Jen2bTyQMNATSQ3/uX8f0wHCwhPZLKxKobqjGIiwkRKl59BSKxly6tAFb7YUWbZxOG7W1nwA6YCIh4SEslt7N2vdOGMuYMS80u72kpIQXX3yREydOkJycTGVlZYBNXl4eixcvBmDNmjVs27aNZcuWsX79eg4fPkxqaipVVVVIKdnzH7/imR8t5ZlF+djtdjRNCzje8uXLWbhwIYsWLeLAgQNkZ2dz+vTpJv1bunQpjz/+OCkpKdhsNn7729+2y3NUiakIJlybdrZlOhl3AXo4XkeHU1fZKCV32ZuWc9Z77aITjMjSkK9AxpM+E/KmQ2yfrvM/DKhqqKJPTB81UbZCcZ84nTUYQgpAx+msaVFMtcbRo0eZP38+ycnJAAGTHAMUFxezZs0aqqqqqK2t5ZFHHgFg+vTpFBQUsGDBAvLy8tAlZEz+Cv+y8WWqbt0gLy8vICrl5uzZs2RkZFBTU0Pfvs1Hqg8fPkxmZiZHjx6lrKyMb3zjG8yYMYM+fUL7LlViKoIxmnaGYc2Ui6Cmk+nuTTsd96Dy0ybScqVGjZMbk8WIJiWNhlF/5o0wJY82ok/d9d8nRKoaqlSKT6FohpYiSG6qq89y9r/+Gl13YDJFMWH8pnZJ9bVEQUEB+/fvJyMjg6KiIo4dOwbA5s2bOXXqFAcPHmTKlCkcP3mKufPm82juDN579whz586lsLCQ3Nxcz7EqKiqYM2cOFRUVxMbGsnv3bmw2G5mZmezdu5dRo0b5nftXv/oVq1atQghBeno6I0aM4JNPPiE7Ozuka1JiKoIJ2zRfG2rKPU07IznPp2tQXR44Us5aZqz3Tcv1HmyIpHF/6S+Y+g4Fc1SXXUKkUt1QrcSUQhECiYmTmZy1s91qpnJzc5k3bx4rVqwgKSmJysrKgOiUzWZj8ODBOBwOdu3aRWpqKgBlZWXk5OSQk5PDoUOHKC8vp8phYcbkcWSO/wnXrl3j/PnzfmJqwIABnDt3jscee4yioiJeffVVsrOzmTt3bpP+DR06lHfffZcZM2Zw8+ZNLl68yMiRI0O6ZlBiKqIJ26adbWmNQBuHa3QVUkKdtVFKzvVX+Slodq9tTB8jDTf0YUhaaCwnj4Z+oyBG1fa0J1UNVaQkpHS1GwpFRJOYOLndolHjx49n9erVzJo1C7PZTFZWFkVFRX42GzZsICcnh/79+5OTk4PNZszNuXLlSkpLS5FSMnv2bCZMnMQL61/ibxc/RUx0FIMGDeL5558POKemaVitVpKTkzlx4gQrVqxo1r8XXniBgoICJk6ciJSSl19+2ZOSDAUlpiIYo2YqDNN8bemAHm5pPvtdQxw1HilnLTU6grsxRRm9l5JGw+g53ghTUrrRqylcrqebU91QzbikcV3thkKh8CE/P5/8/Hy/dWvXrvUsL1myhCVLlgTst2/fPr/31XV2nv7xcn6+/qf0imr+WWc2mzl58iQAR44cadG3lJSUVm3uByWmIpiwbdrp0UdBCIquKEDXnFB9rVFKzvVX87m/bZ9UQyBN+CsfwTQKEoeCWX18uhqV5lMouo67DU7uNjiJj7EQH9P+34f3HMbzocGhtSimwgH1NIhgjKad4SemPHz4BkzKhrQWCvvcyqv0CJSntWzbEuWn4cpxGD7DOIaUcPdWo5RcmRFhqvwMdId339hEI8I0fIZLMLm7fo+E6Pj780fR4dQ766nX6lXDToWiDUgpkdJIIEgpXa9GeYaUjZZbsG9waty22ZFIBIKkhGiiLe33PLI7day1DQBcv3OPKLOpQwRbexG+nilaRQ/XNN+NYsAEJ1+FM7dgSoFRYN0EsuwTY+Hjt+BXb7Zo2yxV1+BMkVEILkxG5Mh2AxpqvDbmaKNmKXkMPDjXm5JLSoe4JJWWi0BUw05FV6LpEoemu/4kTk3H3mjZqUnPdoem49R17E6JU3ft55Q4dB2HU8epy6b30XTsrmP6rjeOJ7G79nVv+/uHEzDdqEFKV8ODJsRQeyOR3HYJn45ASumJgIUr4euZolWkrnV9AbrmhFsX4PoHcP1DKD+NvJ4KPItAB90JH/yy+f1rzcBAELJ122CQmuHTpG97U3JJ6ZCYBuEoPBX3jZqXL/KRUrpEiVdUtCwWvGKlsbDwio9GokQ3RItbwPiJmYB9As/vcLrEjNs/ly+yI1SJCyEgymwiyiSIspiwmExEmwUWs4koszC2uZYtZhO9osz0jrUQZTYZ76MtCNdxBMJ49V3GKMPwLgdrJzz29Q6da5V1SCkRQjAsKa5dU3H3HBpXrd7jh7OQAiWmIhpd64KmnbUVhmi6ftp4/fwsOO4a2+KSjRSbZSp8BEKYwBIL3/0dDJna9PE+OARvvxCcbXNc/xB+vQA0hxGByiu8/3ShImJQ8/J5kVJ6BYc72tGUsGgmcuIvPqQrUtJIWLiO1fRxvZGXps7vaMrWta0jMblFiY/wiDabsDQSJFFmExaT8cC2mALFSpTZV8z47GMWxvFcoifKZCLKIrCYXPv4LAfs04Qwci+bQxjdfOHCBYb2i2vHf8WmibaYGZEc32E1U73Npg49fnsT3t4pWqTDJzp22uHGf7uiTq6/qqvGNpMFBk2CrIVGZ+4hU42mk0IgT3wJH12Ar/4YJuS0LGxSDeFkc2YS9ZWFxI2c1XY/R86C/AP+NVOKbk9HpPl03RACzT34/cSCT2rG0SjV0lp6xve4jQVMsymd5oSRS/R0JGaTMB72JpMrUiJcYsHkIz68YiEhyhIoYFyCw198CE80xVdMuAWHv/hodB6Tj1ixmJr0LxRRomidjhY5kSCi3ESGl4omke05mk9KYySbO113/QP44hxorjx4n1RDMGUvhiHZMHgSRPVq7mDGS/YPIKk5G7Bfu8YXa4wOvbYzV6gtfoWhA8YSl5XVdv/TWil0V3Qqurv+wyUSHHpz6ZngxIIn1eOTZrlQa9TbvX70SyyyLqhj+aeOAs/v1DtWlFhMjaMTLlHis2xxR0FMJnpF+woOf4HhL2CaiZyYBNEWk2fZEz1pMkrTRLTGJVBMSpQoQmTt2rUkJCTw3HPPddg5CgsLiY6OJisri9dff53CwsIAmzt37rBo0SLKysqIjY3ljTfeYMKECSGfW4mpCCakpp32OvjyI1e6ziWgbF8a2yyxMDjTEE5p2Ub0KDE16EO31hpBq63Funkzldt3GLJLCKNI0uGg7vQH9yemOpAzV+/w/qdWHh6ZxJRhD3TquZsrcm0cOWmxyLVxFKTJCEmjiIgnUuJKAzUTeWlSwOhGHUxHEmUWRCddx5QEf7pwl2izwysKfFIt0a7oSXyMxU94RAVEO1wCw9R0tKWl9ExLwsg/7SPUHIIKRQdy/Phx1q1bx9tvv83MmTObtPnZz35GZmYmv//97/nkk0/48Y9/zLvvvhvyuZWYilCklMFPdCyl0YjSHXG6/gHcLDYKvsFIzw2f4U3XDZwAlugQnGtmtaZRtXcvt/7139CsVhLnzSPh61/ni7/9W6TDgYiKIi77K3773I+QuV/x0+DUuNugcbfBSa2rf8q58ipefucTnJrEYhZ8f/oIUhJjcbRS5No4PRRMkas74uK7T0cWuQKuSEdTdSH+YiHKbCI2ykTvWIsnvRKQamkc0Wg22uGNvHhSPX4CpukiW7dIsZgMUfKPHxTz5qVenF7zzY79R1IoujkfVt/lRFUtX+2bwNTE0NvB7Nixg40bNyKEYNKkSezcudNv+9atW9myZQt2u5309HR27txJXFwce/bsYd26dZjNZhITE3nvvfcoKSnh+9//Pna7HV3X2bt3b8Bkx5s2bWL79u2UlpZSXFzMlStXSElJ4fjx42zevNnP9uOPP2bVqlUAPPTQQ1y5coWbN28ycODAkK45KDElhHgU+FfADPxSSvnzRttjgB3AFMAKfFtKeSUkzxQtIqVRuGlqaoRafQ18cdY/ZVdnNbZFJ0DqZJj+rCGeUqdCQv929i1wOpm777/PzX/4OQ0XL9JryhQGFhbSa8J4Y+O2bdjeP4XMmExF2hjqvqjhnkPjfHkVPzt0wSNknp09mpH9W56O5dNbtfzru6U4NYnZJHgyeyh946I84qi2wUmtSzD5iqa7DVqrxbAOTbLlvU8D1rtH3kQ38eD3TbW4xUJctMXPzlfMBBa2GpENd2qmqSLX5gpbP6sp4eOqc2T1n8Kk/hkBBbNmU9sjJecqzvHhzQ+ZOnAqmQMy27Rve/NZ9WdYhIVzFee63BeFIhx5ofQ6xbX3WrSxOTU+rq1HB0zAuIRYeluar8WdkNCLDaOHNLu9pKSEF198kRMnTpCcnExlZWWATV5eHosXLwZgzZo1bNu2jWXLlrF+/XoOHz5MamoqVVVGTeTmzZt59tlneeqpp7Db7WiaFnC85cuXs3DhQhYtWsSBAwfIzs7m9OnTTfqXkZHBvn37mDFjBqdPn+bq1atcv36948WUEMIMvAZ8A7gOfCCEeEtK+bGP2dPAHSlluhDiO8DLwLdD8ixEyj+/Rl2UmVc272Hl3wTO5dOYU+9swmw9xmcZT3siGjW3Kqmz3eX/rtrEuK+O5EbGTPZf+E8scZ/xxEMzPF/gVw59ALZ71H/4CVduHmXg7JmkjBnriZAkJ32JTVxslwfQlUMfUHL0BJX9jHqRzz8p4YtT75ASdcs7wq7iAu7wUGXcCPTUXJIfnG6IpwFj29QiQEpJg1Onzq5xz6Fxz+7knl2nzu50vTfW19k16l2vznOVxAB71mylpq+Zkef/k/TSs1T2SeatOc9wasgk7r31JXVvXqfeobmKZ4dC+W1M727HEvcpzrqR6PeGefxwaJKNRy616d/KqUt2vn8VISAh2uIqZjST4CpqTIqP8yzHx1hIiDF7lnu7Xq/fuce6AyU4NJ0os4nXvjuZKcMe8EsPtVbkqksdTWrGq66hSc3z6t7WeJ1Td7q2OQLXuY8hvfZ17nWahubQuFpzle0fb0fTNX5zycz8MfMZED8Ap+7EqTvRpOZZ9n2vSQ2H7kDTNX876aSqvopLdy55mvSNSBxBfFTXNDW967jLp9WGsF18ZDFb52xVgkqhuA9qnBrun5G6631LYqo1jh49yvz58z3z3TWe5BiguLiYNWvWUFVVRW1tLY888ggA06dPp6CggAULFpCXlwfAtGnTeOmll7h+/Tp5eXkBUSk3Z8+eJSMjg5qaGvr2bX5QyqpVq3j22WfJzMxk4sSJZGVlYW6HgVxCtpJHEEJMA9ZKKR9xvf9fAFLKf/CxOeyyOSmEsAA3gP6yhYNPnTpVfvjhhyFfQFO8svln/POYR5EIorHzg/LfMOhObZO2CX1vMHDoBaPBmW6m9OwcbNXJIAT1FsN9MyYetU/iy5qLfGG2AhITgiRLH6LrY7GIoUxIGASAJp0cu/Fb7sZJqu0Cp8lJXUwNoGPCxJCoZHqZYu7ruqTNwl3nKKQwIZ23cDa8Dxj9P4b1EcTGROGI7osjui81pt6UVJqxSzNCwPDkeKLNJjRp1Fq561p0XXqWNSnRNNera73uavgWLL1IYDjjEcKESbOT+dG/0evuNY5My+HkVyZjio0mxuKtZYm2mIhx/dm0Lzlu3YkmNczCzLR+3+aPJRq6rmE2Q97kFAYmRqPrGjpeIeEVKxo3a+r448WbxjFMkm+M60//PlEeO7cgcUp/YeJZp+t+IkXTNWwNdmrqG+gVLYixiCbFT5PrXMsd0yYvNCwmCxZhwWKyYDaZsQjjNcoUhVmY/dZbTMZfRV0Fn9d6p9sZ1mcYQ3o3/wu1I7luu87VGmNkqVmYWZq1lB9M/EGX+KJQhBMXLlxg7NixQdt/WH2Xvzp3GYcuiTIJ3sxMDynV94tf/IIbN27w0ksv+a33LUAfMWIE+/fvJyMjg6KiIo4dO+aZDPnUqVMcPHiQHTt2cObMGZKSkigrK+PgwYP84he/oLCwkNzcXM9xKyoqmDNnDhUVFcTGxmIymbDZbAwePJi9e/cyatSoZn2VUjJixAjOnz9Pnz59/LY19e8ohDgjpWyyd08wab5UoNzn/XUgpzkbKaVTCFENJAG3GznyQ+CHAEOHtrHLdRu49ECcp7DZKc3cTrMwPe2PLe5jNCvT6ZN0G/PddGqpB+pBGJGFmyYbk3t/lYB5tRuVFpkwMSA2jQt33icGiAHibd7t9yin5aBra3wSsEYiuG6fjsWcDU6gzlg/CgylBZ474a6wapOcu8+aWV2YudN3NEfHX2H3jNOAK+zqdP210DBXk07+n3UXUYO86976HPi82V0AMAkT0UkmBCYsJjMfWC2YK82YhAmLsGAymTALM2bhWmeyYBI+60wuO2Ei2hyN2WKmT4yJYYkuO5Nh53713ddvndtONLPOFPq+bj9993Vf48XKi/zde3+HQ3cQZYritdzXyByY6dnnfgqhz1WcY/GRxZ5jvjj9xS6LBjX2ZerANvYmUygUAExNjOfNzPR2q5nKzc1l3rx5rFixgqSkJCorKwOiU26x43A42LVrF6mpxgCnsrIycnJyyMnJ4dChQ5SXl1NdXc3IkSP5yU9+wrVr1zh//ryfmBowYADnzp3jscceo6ioiFdffZXs7Gzmzp3bpH9VVVXExcURHR3NL3/5S2bOnBkgpO6HTi1Al1JuAbaAEZnqqPOMuVNHdH8HTmnGgsawS1E47E82aWt3fEGvScdBSKQ0EXV1INV3bKTExHA3wRhlZsLEQK03x+4e5WLcFRDGuvEJo4mrikGrG05m7yQEhvC61VBOn5Fj+aRa4jDVci/+GpgkZmHiK30m0ddyfzfOecPBrTsjkJiQWgX2e+8a3WhNJgZNTaRXvzse21u2Bo5cuInUJCaz4ImMFNL6xWMRAtGBfT5r75i4eDIKqUtMUuOBu58xr+Al5k14sNV9L925xIaTG3DqTiwmCxumb2Bc0jh/sdBY/DRap0ZLGaQmpPLLOb9s1/qmzAGZbJ2zNSxqpsLJF4Ui0pmaGN8uhecA48ePZ/Xq1cyaNQuz2UxWVpYn6uRmw4YN5OTk0L9/f3JycrDZjIjDypUrKS0tRUrJ7NmzycjI4OWXX2bnzp1ERUUxaNAgnn8+sGxH0zSsVivJycmcOHGCFStWNOvfhQsXyM/PRwjB+PHj2bZtW7tcd7dM84GR6rv0QBxj7tS1WjN19LUXqDZdJqZhuF/N1OnfHeHC2Y9IwdJqzdTNP31Cn+gG6gdrnVIzdSf5ImkZDxJ7bzBp4yeSMiYwrNtVQ/pvfFrNlWMl9K26zNA/m9imVgfhVOCsUCgUkUZb03yKpmlrmi8YMWUBLgGzMZIsHwDflVKW+Nj8GJgopfwbVwF6npRyQUvH7WgxpVAoFApFT0OJqfah3WumXDVQS4HDGK0R3pBSlggh1gMfSinfArYBO4UQl4FK4DshXodCoVAoFApFRBBUzZSU8g/AHxqt+6nPcj0wv31dUygUCoVCoQh/OrAUWaFQKBQKRWfTWvmOomXu599PiSmFQqFQKLoJsbGxWK1WJajuEyklVquV2NjYNu2n5uZTKBQKhaKbMGTIEK5fv86tW7e62pWIJTY2liFD2taQWIkphUKhUCi6CVFRUYwYMaKr3ehxqDSfQqFQKBQKRQgoMaVQKBQKhUIRAkpMKRQKhUKhUIRAqx3QO+zEQtwCrnbwaZJpNNmyIixQ9yX8UPckPFH3JfxQ9yQ86Yz7MkxK2b+pDV0mpjoDIcSHzbV+V3Qd6r6EH+qehCfqvoQf6p6EJ119X1SaT6FQKBQKhSIElJhSKBQKhUKhCIHuLqa2dLUDiiZR9yX8UPckPFH3JfxQ9yQ86dL70q1rphQKhUKhUCg6mu4emVIoFAqFQqHoULqFmBJCPCqEuCiEuCyEWNXE9hghxG9d208JIYZ3vpc9jyDuywohxMdCiPNCiHeFEMO6ws+eRGv3xMfuW0IIKYRQo5Y6mGDuiRBigeuzUiKE+HVn+9gTCeL7a6gQ4o9CiP9yfYfN7Qo/exJCiDeEEBVCiOJmtgshxL+57tl5IcTkzvIt4sWUEMIMvAZ8ExgHPCmEGNfI7GngjpQyHdgEvNy5XvY8grwv/wVMlVJOAt4E/rFzvexZBHlPEEL0Bp4FTnWuhz2PYO6JEGI08L+A6VLK8cD/7HRHexhBflbWAL+TUmYB3wH+vXO97JEUAY+2sP2bwGjX3w+B1zvBJ6AbiCkgG7gspfxUSmkHdgNPNLJ5AtjuWn4TmC2EEJ3oY0+k1fsipfyjlLLO9fZ9oG3TdCvaSjCfFYANGD846jvTuR5KMPdkMfCalPIOgJSyopN97IkEc18k0Me1nAh80Yn+9UiklO8BlS2YPAHskAbvA32FEIM7w7fuIKZSgXKf99dd65q0kVI6gWogqVO867kEc198eRo41KEeKVq9J66weJqU8mBnOtaDCeZzMgYYI4T4TyHE+0KIln6ZK9qHYO7LWmChEOI68AdgWee4pmiBtj532g1LZ5xEoWgJIcRCYCowq6t96ckIIUzAPwMFXeyKwh8LRtriaxjR2/eEEBOllFVd6pXiSaBISvlPQohpwE4hxAQppd7Vjik6n+4QmfocSPN5P8S1rkkbIYQFIyRr7RTvei7B3BeEEF8HVgOPSykbOsm3nkpr96Q3MAE4JoS4AjwMvKWK0DuUYD4n14G3pJQOKeVnwCUMcaXoOIK5L08DvwOQUp4EYjHmh1N0HUE9dzqC7iCmPgBGCyFGCCGiMQoB32pk8xaQ71r+K+CoVA22OppW74sQIgsoxBBSqg6k42nxnkgpq6WUyVLK4VLK4Rh1bI9LKT/sGnd7BMF8f+3HiEohhEjGSPt92plO9kCCuS/XgNkAQoixGGLqVqd6qWjMW8D3XKP6HgaqpZRfdsaJIz7NJ6V0CiGWAocBM/CGlLJECLEe+FBK+RawDSMEexmjeO07XedxzyDI+/IKkADscY0HuCalfLzLnO7mBHlPFJ1IkPfkMDBHCPExoAErpZQqst6BBHlf/hbYKoRYjlGMXqB+pHcsQojfYPywSHbVqv1vIApASrkZo3ZtLnAZqAO+32m+qXuvUCgUCoVCcf90hzSfQqFQKBQKRZehxJRCoVAoFApFCCgxpVAoFAqFQhECSkwpFAqFQqFQhIASUwqFQqFQKBQhoMSUQqFQKBQKRQgoMaVQKBQKhUIRAkpMKRQKhUKhUITA/wfoQ5BFIJ7ufwAAAABJRU5ErkJggg==\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "DCNN_Predict\n"
      ],
      "metadata": {
        "id": "UJBFXWGtBYZr"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "y_pred_dcnn = di.predict(X_test)\n",
        "y_pred_classes_dcnn = [np.argmax(element) for element in y_pred_dcnn]\n",
        "y_pred_classes_dcnn[:5]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "A9FjzotbBdXF",
        "outputId": "26f03e49-8497-4a59-c399-e45c2358ee5c"
      },
      "execution_count": 575,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[2, 0, 0, 2, 0]"
            ]
          },
          "metadata": {},
          "execution_count": 575
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "predictions_dcnn=di.predict(X_test)\n",
        "predictions_dcnn"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Bs53dNeQIQe1",
        "outputId": "a64ec650-a7b4-4f97-b81a-4b1d7c04853b"
      },
      "execution_count": 576,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[ 0.0,  0.0,  1.0, ...,  0.0,  0.0,  0.0],\n",
              "       [ 1.0,  0.0,  0.0, ...,  0.0,  0.0,  0.0],\n",
              "       [ 1.0,  0.0,  0.0, ...,  0.0,  0.0,  0.0],\n",
              "       ...,\n",
              "       [ 0.0,  0.0,  1.0, ...,  0.0,  0.0,  0.0],\n",
              "       [ 0.9,  0.0,  0.0, ...,  0.0,  0.0,  0.0],\n",
              "       [ 0.0,  0.0,  1.0, ...,  0.0,  0.0,  0.0]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 576
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_predicted_labels_dcnn = [np.argmax(i) for i in y_pred_dcnn]"
      ],
      "metadata": {
        "id": "Znr4CQnjBdt_"
      },
      "execution_count": 577,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "score=tf.nn.softmax(predictions_dcnn[0])\n",
        "score\n",
        "np.argmax(score)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "huRJSw1WBdxj",
        "outputId": "7fd3c180-200a-4b4f-cca7-b70366410e03"
      },
      "execution_count": 578,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "2"
            ]
          },
          "metadata": {},
          "execution_count": 578
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "DCNN_ConfutionMatrix"
      ],
      "metadata": {
        "id": "QaIHdRT1Bwbe"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "cm = tf.math.confusion_matrix(labels=y_test,predictions=y_predicted_labels_dcnn)\n",
        "print(cm)\n",
        "accuracy_score(y_test, y_predicted_labels_dcnn)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Eu3H-J2GB1My",
        "outputId": "e77eef54-1c05-4d05-e442-98c15c068985"
      },
      "execution_count": 579,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tf.Tensor(\n",
            "[[232   1   8   2   0   1   0   0   0   0]\n",
            " [  1  41   9   1   0   2   0   0   0   0]\n",
            " [ 13   4 647   3   1   0   0   0   0   0]\n",
            " [  0   0   3  38   1   0   0   0   0   0]\n",
            " [  1   0   2   2  23   0   2   0   0   0]\n",
            " [  1   1   2   0   1  93   0   0   0   0]\n",
            " [  0   0   0   1   0   0  65   0   1   0]\n",
            " [  0   0   0   0   0   0   0  24   0   0]\n",
            " [  0   0   0   0   0   0   0   0  18   0]\n",
            " [  1   0   0   0   0   0   0   0   0  24]], shape=(10, 10), dtype=int32)\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9488188976377953"
            ]
          },
          "metadata": {},
          "execution_count": 579
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "DCNN_ROC\n"
      ],
      "metadata": {
        "id": "Bb7piUzeB4IL"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#ROC Model\n",
        "\n",
        "y_pred1 = di.predict(X_test)\n",
        "y_pred1=[np.argmax(i) for i in y_pred1]\n",
        "\n",
        "plt.figure(figsize=(10,5))\n",
        "\n",
        "\n",
        "for i in range(10):\n",
        "    nn_fpr_keras, nn_tpr_keras, nn_thresholds_keras = roc_curve(y_test  , y_pred1, pos_label=i)\n",
        "    auc_keras = auc(nn_fpr_keras, nn_tpr_keras)\n",
        "    plt.plot(nn_fpr_keras, nn_tpr_keras,marker='.'.format(), label='class # %0.f' % i)\n",
        "    \n",
        "plt.legend(loc=\"lower right\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 338
        },
        "id": "44TKzeYeBrVL",
        "outputId": "3405ded2-7ac6-45b2-b2a9-48be404c6111"
      },
      "execution_count": 580,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7fcb255682d0>"
            ]
          },
          "metadata": {},
          "execution_count": 580
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# HDC_Predict\n"
      ],
      "metadata": {
        "id": "6fqF4e0pCGuE"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "y_pred_hdc = hdc.predict(X_test)\n",
        "y_pred_classes_hdc = [np.argmax(element) for element in y_pred_hdc]\n",
        "y_pred_classes_hdc[:5]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Kce-b-H9CGuE",
        "outputId": "6e258732-e237-472a-9a2c-61699f88ce3b"
      },
      "execution_count": 581,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[2, 0, 0, 2, 0]"
            ]
          },
          "metadata": {},
          "execution_count": 581
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "predictions_hdc=hdc.predict(X_test)\n",
        "predictions_hdc"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "V4exiyNLcvTB",
        "outputId": "20e74e86-038f-491f-9b7f-bc239321f905"
      },
      "execution_count": 582,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[ 0.0,  0.0,  0.9, ...,  0.0,  0.0,  0.0],\n",
              "       [ 1.0,  0.0,  0.0, ...,  0.0,  0.0,  0.0],\n",
              "       [ 1.0,  0.0,  0.0, ...,  0.0,  0.0,  0.0],\n",
              "       ...,\n",
              "       [ 0.0,  0.0,  0.9, ...,  0.0,  0.0,  0.0],\n",
              "       [ 0.9,  0.0,  0.1, ...,  0.0,  0.0,  0.0],\n",
              "       [ 0.0,  0.0,  1.0, ...,  0.0,  0.0,  0.0]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 582
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_predicted_labels_hdc = [np.argmax(i) for i in y_pred_hdc]"
      ],
      "metadata": {
        "id": "XJJb2WhzCGuE"
      },
      "execution_count": 583,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "score=tf.nn.softmax(predictions_hdc[0])\n",
        "score\n",
        "np.argmax(score)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pGGCBJaWCGuF",
        "outputId": "def6da75-bbfe-4378-fc4a-a26e66d05bb9"
      },
      "execution_count": 584,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "2"
            ]
          },
          "metadata": {},
          "execution_count": 584
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "hdc_ConfutionMatrix"
      ],
      "metadata": {
        "id": "Pn7m85wsCGuF"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "cm = tf.math.confusion_matrix(labels=y_test,predictions=y_predicted_labels_hdc)\n",
        "print(cm)\n",
        "accuracy_score(y_test, y_predicted_labels_hdc)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wvX1pzi3CGuF",
        "outputId": "89262ca1-b955-43eb-e723-bac0db0953e5"
      },
      "execution_count": 585,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tf.Tensor(\n",
            "[[238   0   5   1   0   0   0   0   0   0]\n",
            " [  0  49   5   0   0   0   0   0   0   0]\n",
            " [  1   0 666   0   1   0   0   0   0   0]\n",
            " [  0   0   3  38   1   0   0   0   0   0]\n",
            " [  0   0   1   0  29   0   0   0   0   0]\n",
            " [  1   1   0   0   0  96   0   0   0   0]\n",
            " [  0   0   1   0   0   0  66   0   0   0]\n",
            " [  1   0   0   0   0   0   0  23   0   0]\n",
            " [  0   0   1   0   0   0   0   0  17   0]\n",
            " [  0   0   0   0   0   0   0   0   0  25]], shape=(10, 10), dtype=int32)\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9818897637795275"
            ]
          },
          "metadata": {},
          "execution_count": 585
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "hdc_ROC\n"
      ],
      "metadata": {
        "id": "fbCl2PCSCGuF"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#ROC Model\n",
        "\n",
        "y_pred1 = hdc.predict(X_test)\n",
        "y_pred1=[np.argmax(i) for i in y_pred1]\n",
        "\n",
        "plt.figure(figsize=(10,5))\n",
        "\n",
        "\n",
        "for i in range(10):\n",
        "    nn_fpr_keras, nn_tpr_keras, nn_thresholds_keras = roc_curve(y_test  , y_pred1, pos_label=i)\n",
        "    auc_keras = auc(nn_fpr_keras, nn_tpr_keras)\n",
        "    plt.plot(nn_fpr_keras, nn_tpr_keras,marker='.'.format(), label='class # %0.f' % i)\n",
        "    \n",
        "plt.legend(loc=\"lower right\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 338
        },
        "id": "rpaDux4cCGuF",
        "outputId": "b0596628-5b73-498c-cae3-af28db94b95c"
      },
      "execution_count": 586,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7fcb27b51210>"
            ]
          },
          "metadata": {},
          "execution_count": 586
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#loss curve\n",
        "\n",
        "\n",
        "#plt.plot()\n",
        "#plt.plot(history3.history['loss'])\n",
        "plt.title('loss curve')\n",
        "plt.ylabel('loss')\n",
        "plt.xlabel('epoch')\n",
        "\n",
        "for i in range(3):\n",
        "\n",
        "    if(i==0) :\n",
        "             plt.plot(history1.history['loss'],marker='.'.format(), label='Model # %0.f' % 1)\n",
        "    elif(i==1):\n",
        "             plt.plot(history2.history['loss'],marker='.'.format(), label='Model # %0.f' % 2)\n",
        "    else :\n",
        "             plt.plot(history3.history['loss'],marker='.'.format(), label='Model # %0.f' % 3)\n",
        "plt.legend(loc=\"lower right\")\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 313
        },
        "id": "YlALIUvalyBz",
        "outputId": "c7196c64-c3d1-4f98-b831-61f13de7ca74"
      },
      "execution_count": 587,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7fcb1bd3d550>"
            ]
          },
          "metadata": {},
          "execution_count": 587
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Accuracy curve\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "plt.title('accuracy curve')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "\n",
        "for i in range(3):\n",
        "    if(i==0) :\n",
        "             plt.plot(history1.history['accuracy'],marker='.'.format(), label='Model # %0.f' % 1)\n",
        "    elif(i==1):\n",
        "             plt.plot(history2.history['accuracy'],marker='.'.format(), label='Model # %0.f' % 2)\n",
        "    else :\n",
        "             plt.plot(history3.history['accuracy'],marker='.'.format(), label='Model # %0.f' % 3)\n",
        "plt.legend(loc=\"lower right\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 313
        },
        "id": "3SgEgS4G2R3_",
        "outputId": "e73c12d3-27ad-4228-de03-d4da1d07b0f2"
      },
      "execution_count": 588,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7fcb1bf222d0>"
            ]
          },
          "metadata": {},
          "execution_count": 588
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Validation_Loss_curve\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "plt.title('Validation_lose_curve')\n",
        "plt.ylabel('loss_V')\n",
        "plt.xlabel('epoch')\n",
        "\n",
        "for i in range(3):\n",
        "    if(i==0) :\n",
        "             plt.plot(history1.history['val_loss'],marker='.'.format(), label='Model # %0.f' % 1)\n",
        "    elif(i==1):\n",
        "             plt.plot(history2.history['val_loss'],marker='.'.format(), label='Model # %0.f' % 2)\n",
        "    else :\n",
        "             plt.plot(history3.history['val_loss'],marker='.'.format(), label='Model # %0.f' % 3)\n",
        "plt.legend(loc=\"lower right\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 313
        },
        "id": "DaUCiKDSteMa",
        "outputId": "1447493b-85f4-4d85-ac2a-a91e6cf6daff"
      },
      "execution_count": 589,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7fcb1ade2ed0>"
            ]
          },
          "metadata": {},
          "execution_count": 589
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Compare 3 models in training accuracy\n",
        "\n",
        "print(\"\\t\\t\\t CNN \\t\\t  DCNN   \\t\\t  HDC\")\n",
        "print(f\"Tranning_Accuracy: {train_acc1} {train_acc2} {train_acc3}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4BP8gB2qlUxN",
        "outputId": "63d26b8d-041f-4cd9-9979-320d1b625713"
      },
      "execution_count": 590,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\t\t\t CNN \t\t  DCNN   \t\t  HDC\n",
            "Tranning_Accuracy: 0.9938454031944275 0.9965534210205078 0.9963072538375854\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Compare 3 models in testing accuracy\n",
        "\n",
        "print(\"\\t\\t\\t CNN \\t\\t  DCNN   \\t\\t  HDC\")\n",
        "print(f\"Testing_Accuracy: {test_acc1} {test_acc2} {test_acc3}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tEddM3C2pxyq",
        "outputId": "ac1ad160-72d2-4c3c-aa03-be2f12b44806"
      },
      "execution_count": 591,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\t\t\t CNN \t\t  DCNN   \t\t  HDC\n",
            "Testing_Accuracy: 0.9543306827545166 0.9488189220428467 0.9716535210609436\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Compare 3 models in Required Time\n",
        "\n",
        "print(\"\\t\\t CNN \\t\\t  DCNN   \\t     HDC \")\n",
        "print(f\"Timing:  {time1} {time2} {time3}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "B-iBfQhNpx5o",
        "outputId": "1bbd0725-a635-46a0-9e7f-17fe13bc3e14"
      },
      "execution_count": 592,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\t\t CNN \t\t  DCNN   \t     HDC \n",
            "Timing:  201.97470712661743 206.64966583251953 349.05693340301514\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "CNN_BlockDigram"
      ],
      "metadata": {
        "id": "6-4X720NNBFU"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "tf.keras.utils.plot_model(\n",
        "    cnn,\n",
        "    to_file='model.png',\n",
        "    show_shapes=False,\n",
        "    show_dtype=False,\n",
        "    show_layer_names=True,\n",
        "    rankdir='TB',\n",
        "    expand_nested=False,\n",
        "    dpi=96,\n",
        "    layer_range=None,\n",
        "    show_layer_activations=False\n",
        ")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "q_n_GQaTK6Ke",
        "outputId": "6c115abf-76e8-4941-811c-8c42501ffe5b"
      },
      "execution_count": 596,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<IPython.core.display.Image object>"
            ],
            "image/png": "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\n"
          },
          "metadata": {},
          "execution_count": 596
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "DNN_BlockDigram"
      ],
      "metadata": {
        "id": "BfiK6gGMNHv6"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "tf.keras.utils.plot_model(\n",
        "    di,\n",
        "    to_file='model.png',\n",
        "    show_shapes=False,\n",
        "    show_dtype=False,\n",
        "    show_layer_names=True,\n",
        "    rankdir='TB',\n",
        "    expand_nested=False,\n",
        "    dpi=96,\n",
        "    layer_range=None,\n",
        "    show_layer_activations=False\n",
        ")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "43cfed7b-8160-4785-fd95-3a1e0b022609",
        "id": "oZep7h3yMfOy"
      },
      "execution_count": 597,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<IPython.core.display.Image object>"
            ],
            "image/png": "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\n"
          },
          "metadata": {},
          "execution_count": 597
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "HDC_BlockDigram"
      ],
      "metadata": {
        "id": "ljnVU64RNSfV"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "tf.keras.utils.plot_model(\n",
        "    hdc,\n",
        "    to_file='model.png',\n",
        "    show_shapes=False,\n",
        "    show_dtype=False,\n",
        "    show_layer_names=True,\n",
        "    rankdir='TB',\n",
        "    expand_nested=False,\n",
        "    dpi=96,\n",
        "    layer_range=None,\n",
        "    show_layer_activations=False\n",
        ")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "531cda78-7215-47a6-a4c6-758ec8788883",
        "id": "Wtju2H12MuZ3"
      },
      "execution_count": 598,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<IPython.core.display.Image object>"
            ],
            "image/png": "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\n"
          },
          "metadata": {},
          "execution_count": 598
        }
      ]
    }
  ]
}