Bee_Classification.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "name": "notebook.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "machine_shape": "hm"
    },
    "kernelspec": {
      "display_name": "tensorflow",
      "language": "python",
      "name": "tensorflow"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.7.9"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DdS2aNTgvHdr"
      },
      "source": [
        "# Final Project - Classification of Honey Bee and Bumble Bee"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RXdPo8ibvHdr"
      },
      "source": [
        "In this project we are implementing a deep learning model that automatically detects honey bess and bumble bees. Once the model is trained, we evaluate it on the test data to see how good the model can perform the defined task.\n",
        "\n",
        "<em>A honey bee.</em></p>\n",
        "<p><img src=\"https://assets.datacamp.com/production/project_555/img/92_notebook.jpg\" alt=\"honey bee\">\n",
        "\n",
        "<em>A bumble bee .</em></p>\n",
        "\n",
        "<p><img src=\"https://assets.datacamp.com/production/project_555/img/20_notebook.jpg\" alt=\"bumble bee\">\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "z6EVl-L1vHds"
      },
      "source": [
        "## Mounting the drive"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nJotNoJEM9uf",
        "outputId": "d8454acf-650e-4ac8-ae8c-d1b821e55ce1"
      },
      "source": [
        "#mount drive\n",
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sPTCwJgJvHdt"
      },
      "source": [
        "## 1. Import the necessary libraries"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "dc": {
          "key": "3"
        },
        "id": "gPB9mUBeM5K7",
        "tags": [
          "sample_code"
        ],
        "outputId": "d137b1a0-9c23-4ca8-9e87-c2873019f23d"
      },
      "source": [
        "import pickle\n",
        "from pathlib import Path\n",
        "from skimage import io\n",
        "\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline\n",
        "\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import classification_report\n",
        "\n",
        "# import keras library\n",
        "import keras\n",
        "\n",
        "# import Sequential from the keras models module\n",
        "from keras.models import Sequential\n",
        "\n",
        "# import Dense, Dropout, Flatten, Conv2D, MaxPooling2D from the keras layers module\n",
        "from keras.layers import Dense, Dropout, Flatten, Conv2D, MaxPooling2D"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Using TensorFlow backend.\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "dc": {
          "key": "10"
        },
        "deletable": false,
        "editable": false,
        "id": "EovbZ1j6M5K8",
        "run_control": {
          "frozen": true
        },
        "tags": [
          "context"
        ]
      },
      "source": [
        "## 2. Load image labels\n",
        "Now lets look at the labels for our data. We will load our <code>labels.csv</code> file into a DataFrame called <code>labels</code>, where the index column is the image name and the <code>genus</code> column tells us the bee type. <code>genus</code> takes the value of either <code>0.0</code> (honey bee) or <code>1.0</code> (bumble bee)."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "dc": {
          "key": "10"
        },
        "id": "im7qF3q2M5K8",
        "tags": [
          "sample_code"
        ],
        "outputId": "baa316b1-aee4-4bd6-ae3a-b0e91b0f9215"
      },
      "source": [
        "# load labels.csv from datasets folder using pandas\n",
        "labels = pd.read_csv('/content/drive/MyDrive/Naive-Bees-Deep-Learning-with-Images/datasets/labels.csv', index_col=0)\n",
        "\n",
        "# print value counts for genus\n",
        "print(labels.genus.count())\n",
        "DEADLINE: 16.02.2020\n",
        "\n",
        "# assign the genus label values to y\n",
        "y = labels.genus.values"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "3968\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "dc": {
          "key": "17"
        },
        "deletable": false,
        "editable": false,
        "id": "4OgQMlCPM5K-",
        "run_control": {
          "frozen": true
        },
        "tags": [
          "context"
        ]
      },
      "source": [
        "Let's look at the data for a sample image. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 320
        },
        "dc": {
          "key": "17"
        },
        "id": "ARjVdxxEM5K_",
        "tags": [
          "sample_code"
        ],
        "outputId": "088fe0d5-66e6-42b7-8b59-21ca6e939eca"
      },
      "source": [
        "# load an image and explore\n",
        "%cd /content/drive/MyDrive/Naive-Bees-Deep-Learning-with-Images/\n",
        "example_image = io.imread('datasets/images/{}.jpg'.format(labels.index[0]))\n",
        "\n",
        "# show image\n",
        "plt.imshow(example_image)\n",
        "\n",
        "# print shape\n",
        "print('Example image has shape: ', example_image.shape)\n",
        "                    \n",
        "# print color channel values for top left pixel\n",
        "print('RGB values for the top left pixel are:', example_image[0,0,:])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content/drive/MyDrive/Naive-Bees-Deep-Learning-with-Images\n",
            "Example image has shape:  (200, 200, 3)\n",
            "RGB values for the top left pixel are: [129 109 102]\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "dc": {
          "key": "24"
        },
        "deletable": false,
        "editable": false,
        "id": "6blMwhk3M5LA",
        "run_control": {
          "frozen": true
        },
        "tags": [
          "context"
        ]
      },
      "source": [
        "## 4. Normalize image data\n",
        "Now we need to normalize the image data. Normalization is a general term that means changing the scale of our data so it is consistent.</p>\n",
        "<p>In this case, we want each feature to have a similar range so our neural network can learn effectively across all the features.\n",
        "<p>We will scale our data so that it has a mean of 0 and standard deviation of 1. We'll use sklearn's <code>StandardScaler</code> to do this, which entails taking each value, subtracting the mean, and then dividing by the standard deviation. We need to do this for each color channel (i.e. each feature) individually."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "dc": {
          "key": "24"
        },
        "id": "ng74cpX_M5LA",
        "tags": [
          "sample_code"
        ],
        "outputId": "f140152d-c4af-4344-e7d5-2ce12e39c879"
      },
      "source": [
        "%cd /content/drive/MyDrive/Naive-Bees-Deep-Learning-with-Images/\n",
        "# initialize standard scaler\n",
        "ss = StandardScaler()\n",
        "count = 0\n",
        "image_list = []\n",
        "for i in labels.index:\n",
        "    # load image\n",
        "    img = io.imread('datasets/images/{}.jpg'.format(i)).astype(np.float64)\n",
        "    \n",
        "    # for each channel, apply standard scaler's fit_transform method\n",
        "    for channel in range(img.shape[2]):\n",
        "        img[:, :, channel] = ss.fit_transform(img[:, :, channel])\n",
        "        # if(img.shape != (200, 200, 3)):\n",
        "        #   count = count+1\n",
        "        #   print(\"datasets/images/{}.jpg\".format(i) + \": \" +str(img.shape))\n",
        "\n",
        "    # append to list of all images\n",
        "    image_list.append(img)\n",
        "\n",
        "# print(count)\n",
        "# convert image list to single array\n",
        "X = np.array(image_list)\n",
        "\n",
        "# print shape of X\n",
        "print(X.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content/drive/MyDrive/Naive-Bees-Deep-Learning-with-Images\n",
            "(3968, 200, 200, 3)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "dc": {
          "key": "31"
        },
        "deletable": false,
        "editable": false,
        "id": "cBV5K0XzM5LB",
        "run_control": {
          "frozen": true
        },
        "tags": [
          "context"
        ]
      },
      "source": [
        "## 5. Split into train, test, and evaluation sets\n",
        "Now we split the data into train, test, and evaluation sets. To do this, we'll first allocate 20% of the data into our evaluation, or holdout, set. This is data that the model never sees during training and will be used to score our trained model.</p>\n",
        "<p>We will then split the remaining data, 60/40. We will pass both the train and test sets into the neural network. </p>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "dc": {
          "key": "31"
        },
        "id": "DZKr8Lj9M5LB",
        "tags": [
          "sample_code"
        ],
        "outputId": "6e1a3c64-8684-43a1-c107-1da419f0fc84"
      },
      "source": [
        "# split out evaluation sets (x_eval and y_eval)\n",
        "x_interim, x_eval, y_interim, y_eval = train_test_split(X,\n",
        "                                           y,\n",
        "                                           test_size=0.2,\n",
        "                                           random_state=52)\n",
        "\n",
        "# split remaining data into train and test sets\n",
        "x_train, x_test, y_train, y_test = train_test_split(x_interim,\n",
        "                                           y_interim,\n",
        "                                           test_size=0.4,\n",
        "                                           random_state=52)\n",
        "\n",
        "# examine number of samples in train, test, and validation sets\n",
        "print('x_train shape:', x_train.shape)\n",
        "print(x_train.shape[0], 'train samples')\n",
        "print(x_test.shape[0], 'test samples')\n",
        "print(x_eval.shape[0], 'eval samples')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "x_train shape: (1904, 200, 200, 3)\n",
            "1904 train samples\n",
            "1270 test samples\n",
            "794 eval samples\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "dc": {
          "key": "17"
        },
        "deletable": false,
        "editable": false,
        "run_control": {
          "frozen": true
        },
        "tags": [
          "context"
        ],
        "id": "al06iZGY2EuH"
      },
      "source": [
        "This is the number of images which we have in our dataset.\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "dc": {
          "key": "38"
        },
        "deletable": false,
        "editable": false,
        "id": "3kBzpsojM5LC",
        "run_control": {
          "frozen": true
        },
        "tags": [
          "context"
        ]
      },
      "source": [
        "## 6. Model building\n",
        "<p>Now we build the deep learning model, a convolutional neural network (CNN). </p>\n",
        "<p>Building a CNN begins with specifying the model type. In our case, we'll use a Sequential model, which is a linear stack of layers. We'll then add two convolutional layers. Then we add a Max Pooling layer, Conv2D, Droupout to avoid overfitting with 25% drop, Flatten, Dense, and another dropout layer to safeguard against overfitting, this time with a rate of 50% drop. And now a final Dense layer which calculates the probability the image is either a bumble bee or honey bee.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "collapsed": true,
        "dc": {
          "key": "38"
        },
        "id": "h4zc7MHLM5LD",
        "tags": [
          "sample_code"
        ]
      },
      "source": [
        "# set model constants\n",
        "num_classes = 1\n",
        "\n",
        "# define model as Sequential\n",
        "model = Sequential()\n",
        "\n",
        "# first convolutional layer with 32 filters\n",
        "model.add(Conv2D(32, kernel_size=(3, 3), activation='relu', input_shape=(200, 200, 3)))\n",
        "\n",
        "# add a second 2D convolutional layer with 64 filters\n",
        "model.add(Conv2D(64, kernel_size=(3, 3), activation='relu', input_shape=(200, 200, 3)))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "dc": {
          "key": "45"
        },
        "id": "Xr3EF2TQM5LE",
        "tags": [
          "sample_code"
        ],
        "outputId": "8ca953ca-01d6-4f33-a7f1-2478db871030"
      },
      "source": [
        "# reduce dimensionality through max pooling\n",
        "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "# third convolutional layer with 64 filters\n",
        "model.add(Conv2D(64, kernel_size=(3, 3), activation='relu'))\n",
        "# add dropout to prevent over fitting\n",
        "model.add(Dropout(0.25))\n",
        "# necessary flatten step preceeding dense layer\n",
        "model.add(Flatten())\n",
        "# fully connected layer\n",
        "model.add(Dense(128, activation='relu'))\n",
        "\n",
        "# add additional dropout to prevent overfitting\n",
        "model.add(Dropout(0.5))\n",
        "\n",
        "# prediction layers\n",
        "model.add(Dense(num_classes, activation='sigmoid', name='preds'))\n",
        "\n",
        "# show model summary\n",
        "model.summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d (Conv2D)              (None, 198, 198, 32)      896       \n",
            "_________________________________________________________________\n",
            "conv2d_1 (Conv2D)            (None, 196, 196, 64)      18496     \n",
            "_________________________________________________________________\n",
            "max_pooling2d (MaxPooling2D) (None, 98, 98, 64)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_2 (Conv2D)            (None, 96, 96, 64)        36928     \n",
            "_________________________________________________________________\n",
            "dropout (Dropout)            (None, 96, 96, 64)        0         \n",
            "_________________________________________________________________\n",
            "flatten (Flatten)            (None, 589824)            0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 128)               75497600  \n",
            "_________________________________________________________________\n",
            "dropout_1 (Dropout)          (None, 128)               0         \n",
            "_________________________________________________________________\n",
            "preds (Dense)                (None, 1)                 129       \n",
            "=================================================================\n",
            "Total params: 75,554,049\n",
            "Trainable params: 75,554,049\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "dc": {
          "key": "52"
        },
        "deletable": false,
        "editable": false,
        "id": "KmqQn5XmM5LF",
        "run_control": {
          "frozen": true
        },
        "tags": [
          "context"
        ]
      },
      "source": [
        "## 8. Compile and train model\n",
        "<p>As the model architecture is defined, now we compile the model for training. For this we need to specify the loss function (what we're trying to minimize), the optimizer (how we want to go about minimizing the loss), and the metric (how we'll judge the performance of the model).</p>\n",
        "<p>Then, we'll call <code>.fit</code></a> to begin the trainig the process. </p>\n",
        "\n",
        "We are using checkpoints to save the best weights by checking the validation loss. We use earlystopping here to avoid overfitting of the model."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "dc": {
          "key": "52"
        },
        "id": "OZ3Xmyo2M5LF",
        "tags": [
          "sample_code"
        ],
        "outputId": "16c59071-a241-4ef2-88d3-5262c32fa52b"
      },
      "source": [
        "from tensorflow.keras.callbacks import ModelCheckpoint, EarlyStopping\n",
        "\n",
        "model.compile(\n",
        "    # set the loss as binary_crossentropy\n",
        "    loss=keras.losses.binary_crossentropy,\n",
        "    # set the optimizer as stochastic gradient descent\n",
        "    optimizer=keras.optimizers.SGD(lr=0.001),\n",
        "    # set the metric as accuracy\n",
        "    metrics=['accuracy']\n",
        ")\n",
        "\n",
        "checkpoint = ModelCheckpoint(\"/content/drive/MyDrive/Naive-Bees-Deep-Learning-with-Images/datasets/checkpoint_notebook/weights-improvement-{epoch:02d}-{val_loss:.2f}.hdf5\", monitor=\"val_loss\", mode=\"min\", save_best_only = True, verbose=1)\n",
        "earlystop = EarlyStopping(monitor = 'val_loss',  min_delta = 0,  patience = 5, verbose = 1, restore_best_weights = True)\n",
        "callbacks = [checkpoint,earlystop]\n",
        "\n",
        "history= model.fit(\n",
        "    x_train,\n",
        "    y_train,\n",
        "    epochs=200,\n",
        "    verbose=1,\n",
        "    validation_data=(x_test, y_test),\n",
        "    callbacks=callbacks\n",
        ")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/200\n",
            "60/60 [==============================] - 7s 80ms/step - loss: 0.5554 - accuracy: 0.7499 - val_loss: 0.5053 - val_accuracy: 0.7874\n",
            "\n",
            "Epoch 00001: val_loss improved from inf to 0.50532, saving model to /content/drive/MyDrive/Naive-Bees-Deep-Learning-with-Images/datasets/checkpoint_notebook/weights-improvement-01-0.51.hdf5\n",
            "Epoch 2/200\n",
            "60/60 [==============================] - 4s 68ms/step - loss: 0.5041 - accuracy: 0.7858 - val_loss: 0.5049 - val_accuracy: 0.7874\n",
            "\n",
            "Epoch 00002: val_loss improved from 0.50532 to 0.50486, saving model to /content/drive/MyDrive/Naive-Bees-Deep-Learning-with-Images/datasets/checkpoint_notebook/weights-improvement-02-0.50.hdf5\n",
            "Epoch 3/200\n",
            "60/60 [==============================] - 4s 68ms/step - loss: 0.4947 - accuracy: 0.7812 - val_loss: 0.4915 - val_accuracy: 0.7874\n",
            "\n",
            "Epoch 00003: val_loss improved from 0.50486 to 0.49149, saving model to /content/drive/MyDrive/Naive-Bees-Deep-Learning-with-Images/datasets/checkpoint_notebook/weights-improvement-03-0.49.hdf5\n",
            "Epoch 4/200\n",
            "60/60 [==============================] - 4s 68ms/step - loss: 0.4311 - accuracy: 0.8045 - val_loss: 0.4847 - val_accuracy: 0.7945\n",
            "\n",
            "Epoch 00004: val_loss improved from 0.49149 to 0.48466, saving model to /content/drive/MyDrive/Naive-Bees-Deep-Learning-with-Images/datasets/checkpoint_notebook/weights-improvement-04-0.48.hdf5\n",
            "Epoch 5/200\n",
            "60/60 [==============================] - 4s 68ms/step - loss: 0.4434 - accuracy: 0.8053 - val_loss: 0.5028 - val_accuracy: 0.7811\n",
            "\n",
            "Epoch 00005: val_loss did not improve from 0.48466\n",
            "Epoch 6/200\n",
            "60/60 [==============================] - 4s 68ms/step - loss: 0.4166 - accuracy: 0.8124 - val_loss: 0.5240 - val_accuracy: 0.7528\n",
            "\n",
            "Epoch 00006: val_loss did not improve from 0.48466\n",
            "Epoch 7/200\n",
            "60/60 [==============================] - 4s 68ms/step - loss: 0.4146 - accuracy: 0.8134 - val_loss: 0.5800 - val_accuracy: 0.6858\n",
            "\n",
            "Epoch 00007: val_loss did not improve from 0.48466\n",
            "Epoch 8/200\n",
            "60/60 [==============================] - 4s 69ms/step - loss: 0.4314 - accuracy: 0.7979 - val_loss: 0.4888 - val_accuracy: 0.7921\n",
            "\n",
            "Epoch 00008: val_loss did not improve from 0.48466\n",
            "Epoch 9/200\n",
            "60/60 [==============================] - 4s 67ms/step - loss: 0.3907 - accuracy: 0.8213 - val_loss: 0.4890 - val_accuracy: 0.7827\n",
            "\n",
            "Epoch 00009: val_loss did not improve from 0.48466\n",
            "Restoring model weights from the end of the best epoch.\n",
            "Epoch 00009: early stopping\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "dc": {
          "key": "59"
        },
        "deletable": false,
        "editable": false,
        "id": "JhuaxqQOM5LG",
        "run_control": {
          "frozen": true
        },
        "tags": [
          "context"
        ]
      },
      "source": [
        "## 9. Visualize model training history\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 530
        },
        "id": "9DIbKd7jkwNU",
        "outputId": "33314317-b972-4cb7-8f9e-8b92fc6e4f86"
      },
      "source": [
        "# summarize history for loss\n",
        "# plt.plot(history.history['loss'], label='train loss')\n",
        "plt.plot(history.history['val_loss'], label='val loss')\n",
        "plt.legend()\n",
        "plt.show()\n",
        "plt.savefig('Val_loss')\n",
        "\n",
        "# plt.plot(history.history['loss'], label='train loss')\n",
        "plt.plot(history.history['val_accuracy'], label='val accuracy')\n",
        "plt.legend()\n",
        "plt.show()\n",
        "plt.savefig('Val_accuracy')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 0 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "dc": {
          "key": "73"
        },
        "deletable": false,
        "editable": false,
        "run_control": {
          "frozen": true
        },
        "tags": [
          "context"
        ],
        "id": "MPrg7dMd09oM"
      },
      "source": [
        "## 9. Load saved model and score\n",
        "<p>Now we'll load the saved and evaluateto see how well the model did at classifying bumble bees and honey bees for the test and validation sets. Given that our classes are balanced, a model that predicts <code>1.0</code> for every image would get an accuracy around <code>0.5</code>.</p>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "dc": {
          "key": "59"
        },
        "id": "tS6AeRENM5LG",
        "tags": [
          "sample_code"
        ],
        "outputId": "1b32fdb3-4c65-4fa6-f726-618ffdc9f9a2"
      },
      "source": [
        "# load pre-trained model\n",
        "pretrained_cnn = keras.models.load_model('/content/drive/MyDrive/Naive-Bees-Deep-Learning-with-Images/datasets/checkpoint_notebook/weights-improvement-04-0.48.hdf5')\n",
        "\n",
        "# evaluate model on test set\n",
        "score = pretrained_cnn.evaluate(x_test, y_test, verbose=0)\n",
        "print('Test loss:', score[0])\n",
        "print('Test accuracy:', score[1])\n",
        "\n",
        "print(\"\")\n",
        "\n",
        "# evaluate model on holdout set\n",
        "eval_score = pretrained_cnn.evaluate(x_eval, y_eval, verbose=0)\n",
        "# print loss score\n",
        "print('Eval loss:', eval_score[0])\n",
        "# print accuracy score\n",
        "print('Eval accuracy:', eval_score[1])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Test loss: 0.4846564829349518\n",
            "Test accuracy: 0.7944881916046143\n",
            "\n",
            "Eval loss: 0.4536413848400116\n",
            "Eval accuracy: 0.8110831379890442\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "dc": {
          "key": "73"
        },
        "deletable": false,
        "editable": false,
        "id": "Ri90oXLvM5LI",
        "run_control": {
          "frozen": true
        },
        "tags": [
          "context"
        ]
      },
      "source": [
        "## 10. Generate predictions\n",
        "<p>Now lets access probabilities and class predictions for individual images using the <code>.predict</code> and <code>.predict_classes</code> methods.</p>\n",
        "<p>As we now have a deep learning model that can be used to identify honey bees and bumble bees in images, lets test this on our evaluation dataset which the model has never seen during training."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "dc": {
          "key": "73"
        },
        "id": "QwGU5dn2M5LI",
        "tags": [
          "sample_code"
        ],
        "outputId": "35a81815-bbfd-46ef-cc97-9a2e10789379"
      },
      "source": [
        "# predicted probabilities for x_eval\n",
        "y_proba = pretrained_cnn.predict(x_eval)\n",
        "\n",
        "print(\"First five probabilities:\")\n",
        "print(y_proba[:5])\n",
        "print(\"\")\n",
        "\n",
        "# predicted classes for x_eval\n",
        "y_pred = pretrained_cnn.predict_classes(x_eval)\n",
        "\n",
        "print(\"First five class predictions:\")\n",
        "print(y_pred[:5])\n",
        "print(\"\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "First five probabilities:\n",
            "[[0.81092304]\n",
            " [0.911364  ]\n",
            " [0.7142421 ]\n",
            " [0.8865571 ]\n",
            " [0.62124336]]\n",
            "\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/sequential.py:450: UserWarning: `model.predict_classes()` is deprecated and will be removed after 2021-01-01. Please use instead:* `np.argmax(model.predict(x), axis=-1)`,   if your model does multi-class classification   (e.g. if it uses a `softmax` last-layer activation).* `(model.predict(x) > 0.5).astype(\"int32\")`,   if your model does binary classification   (e.g. if it uses a `sigmoid` last-layer activation).\n",
            "  warnings.warn('`model.predict_classes()` is deprecated and '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "First five class predictions:\n",
            "[[1]\n",
            " [1]\n",
            " [1]\n",
            " [1]\n",
            " [1]]\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}