DS4B-BasicsOfANN.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "c98662e7",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>\n",
       "h1, h2, h3, h4, h5, h6 {\n",
       "  font-family: Verdana, \"DejaVu Sans\", \"Bitstream Vera Sans\", Geneva, sans-serif;\n",
       "  font-weight: bold;\n",
       "}\n",
       "body, intro, introE, introT, rq, rqE, rqT, ex, exE, app, appE, sol, todo, figure  {\n",
       "  font-family: Verdana, \"DejaVu Sans\", \"Bitstream Vera Sans\", Geneva, sans-serif;\n",
       "  font-weight: 400;\n",
       "  font-size: 12px;\n",
       "}\n",
       "h1 {\n",
       "  border: 0 solid #333;\n",
       "  padding: 30px ;\n",
       "  color: white;\n",
       "  background: #b11d01;\n",
       "  text-align: center;\n",
       "}\n",
       "h2 {\n",
       "  border: 3px solid #333;\n",
       "  padding: 18px ;\n",
       "  color: #b11d01;\n",
       "  background: #ffffff;\n",
       "  text-align: center;\n",
       "}\n",
       "h3 {\n",
       "  border: 0 solid #333;\n",
       "  padding: 12px ;\n",
       "  color: #000000;\n",
       "  background: #c1c1c1;\n",
       "  text-align: left;\n",
       "}\n",
       "h4 {\n",
       "  border: 0 solid #333;\n",
       "  padding: 2px ;\n",
       "  color: #000000;\n",
       "  background: #d9fffc;\n",
       "  text-align: left;\n",
       "}\n",
       "h5 {\n",
       "  border: 1px solid #333;\n",
       "  padding: 2px ;\n",
       "  color: #000000;\n",
       "  background: #ffffff;\n",
       "  text-align: left;\n",
       "}\n",
       ".introT::before {    \n",
       "    content: attr(title);\n",
       "    background-color: #cecece;\n",
       "    margin-left:-0.5em;\n",
       "    margin-right:-0.5em;\n",
       "    padding-left:0.5em;\n",
       "    padding-right:0.5em;\n",
       "    font-weight: 600;\n",
       "    display: block;\n",
       "}\n",
       ".introT {    \n",
       "    background-color: #cecece80;\n",
       "    border-color: #969696;\n",
       "    border-left: 5px solid #969696;\n",
       "    padding: 0.5em;\n",
       "}\n",
       ".intro {    \n",
       "    background-color: #cecece80;\n",
       "    border-color: #969696;\n",
       "    border-left: 5px solid #969696;\n",
       "    padding: 0.5em;\n",
       "}\n",
       ".introE {    \n",
       "    background-color: #cecece80;\n",
       "    border-color: #969696;\n",
       "    border-left: 5px solid #969696;\n",
       "    padding: 0.5em;\n",
       "    color : #117996;\n",
       "}\n",
       ".rq::before {    \n",
       "    background-color: #fcd3d3;\n",
       "    color: #ff0000;\n",
       "    content:\"Remarque\";\n",
       "    margin-left:-0.5em;\n",
       "    margin-right:-0.5em;\n",
       "    padding-left:0.5em;\n",
       "    padding-right:0.5em;\n",
       "    font-weight: 600;\n",
       "    display: block;\n",
       "}\n",
       ".rq {    \n",
       "    background-color: #fcf2f2;\n",
       "    border-color: #dFb5b4;\n",
       "    border-left: 5px solid #dfb5b4;\n",
       "    padding: 0.5em;\n",
       "}\n",
       ".rqE::before {    \n",
       "    background-color: #fcd3d3;\n",
       "    color: #ff0000;\n",
       "    content:\"Pay attention\";\n",
       "    margin-left:-0.5em;\n",
       "    margin-right:-0.5em;\n",
       "    padding-left:0.5em;\n",
       "    padding-right:0.5em;\n",
       "    font-weight: 600;\n",
       "    display: block;\n",
       "}\n",
       ".rqE {    \n",
       "    background-color: #fcd3d380;\n",
       "    border-color: #dFb5b4;\n",
       "    border-left: 5px solid #dfb5b4;\n",
       "    padding: 0.5em;\n",
       "}\n",
       ".rqT::before {    \n",
       "    background-color: #fcd3d3;\n",
       "    color: #ff0000;\n",
       "    content: attr(title);\n",
       "    margin-left:-0.5em;\n",
       "    margin-right:-0.5em;\n",
       "    padding-left:0.5em;\n",
       "    padding-right:0.5em;\n",
       "    font-weight: 600;\n",
       "    display: block;\n",
       "}\n",
       ".rqT {    \n",
       "    background-color: #fcd3d380;\n",
       "    border-color: #dFb5b4;\n",
       "    border-left: 5px solid #dfb5b4;\n",
       "    padding: 0.5em;\n",
       "}\n",
       ".exold {    \n",
       "    background-color: #b2dbea80;\n",
       "    border-color: #0055ff;\n",
       "    border-left: 10px solid #0055ff;\n",
       "    padding: 0.5em;\n",
       "}\n",
       ".ex {    \n",
       "    background-color: #b2dbea80;\n",
       "    padding: 0.5em;\n",
       "    padding-top: 0em;\n",
       "    position:relative;\n",
       "}\n",
       ".ex::before {\n",
       "    background-color: #b2dbea;\n",
       "    content:\"Exercice. \" attr(title);\n",
       "    margin-left:-0.5em;\n",
       "    margin-right:-0.5em;\n",
       "    padding-left:0.5em;\n",
       "    padding-right:0.5em;\n",
       "    font-weight: 600;\n",
       "    display: block;\n",
       "}\n",
       ".exE {    \n",
       "    background-color: #b2dbea80;\n",
       "    padding: 0.5em;\n",
       "    padding-top: 0em;\n",
       "    position:relative;\n",
       "    color : #117996;\n",
       "}\n",
       ".exE::before {\n",
       "    background-color: #b2dbea;\n",
       "    content:\"Exercise. \" attr(title);\n",
       "    margin-left:-0.5em;\n",
       "    margin-right:-0.5em;\n",
       "    padding-left:0.5em;\n",
       "    padding-right:0.5em;\n",
       "    font-weight: 600;\n",
       "    display: block;\n",
       "    color : #117996;\n",
       "}\n",
       ".app {    \n",
       "    background-color: #b2dbea80;\n",
       "    padding: 0.5em;\n",
       "    padding-top: 0em;\n",
       "    position:relative;\n",
       "}\n",
       ".app::before {\n",
       "    background-color: #b2dbea;\n",
       "    content:\"Application\";\n",
       "    margin-left:-0.5em;\n",
       "    margin-right:-0.5em;\n",
       "    padding-left:0.5em;\n",
       "    padding-right:0.5em;\n",
       "    font-weight: 600;\n",
       "    display: block;\n",
       "}\n",
       ".appE {    \n",
       "    background-color: #b2dbea80;\n",
       "    padding: 0.5em;\n",
       "    padding-top: 0em;\n",
       "    color : #117996;\n",
       "    position:relative;\n",
       "}\n",
       ".appE::before {\n",
       "    background-color: #b2dbea;\n",
       "    content:\"Application\";\n",
       "    margin-left:-0.5em;\n",
       "    margin-right:-0.5em;\n",
       "    padding-left:0.5em;\n",
       "    padding-right:0.5em;\n",
       "    font-weight: 600;\n",
       "    color : #117996;\n",
       "    display: block;\n",
       "}\n",
       ".sol {    \n",
       "    background-color: #bbeab880;\n",
       "    padding: 0.5em;\n",
       "    padding-top: 0em;\n",
       "    position:relative;\n",
       "}\n",
       ".sol::before {\n",
       "    background-color: #bbeab8;\n",
       "    content:\"Answer\";\n",
       "    margin-left:-0.5em;\n",
       "    margin-right:-0.5em;\n",
       "    padding-left:0.5em;\n",
       "    padding-right:0.5em;\n",
       "    font-weight: 600;\n",
       "    display: block;\n",
       "}\n",
       ".com {    \n",
       "    background-color: #ffff7f80;\n",
       "    padding: 0.5em;\n",
       "    padding-top: 0em;\n",
       "    position:relative;\n",
       "}\n",
       ".com::before {\n",
       "    background-color: #ffff7f;\n",
       "    content:\"Comment.\";\n",
       "    margin-left:-0.5em;\n",
       "    margin-right:-0.5em;\n",
       "    padding-left:0.5em;\n",
       "    padding-right:0.5em;\n",
       "    font-weight: 600;\n",
       "    display: block;\n",
       "}\n",
       "div.todo:before {\n",
       "    content:url(data:image/png;base64,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);\n",
       "    float:left;\n",
       "    margin-left:20px;\n",
       "    margin-right:20px;\n",
       "    margin-top:20px;\n",
       "    margin-bottom:20px;\n",
       "}\n",
       "div.todo{\n",
       "    font-size: 1.1em;\n",
       "    margin-top:40px;\n",
       "    background-color: #b2dbea80;\n",
       "}\n",
       "div.todo ul{\n",
       "    margin: 0.2em;\n",
       "}\n",
       "div.todo li{\n",
       "    margin-left:60px;\n",
       "    margin-top:0;\n",
       "    margin-bottom:0;\n",
       "}\n",
       "\n",
       "figure {\n",
       "    border: 0px;\n",
       "    text-align: center;\n",
       "    margin: 5px; /* adjust as needed */\n",
       "    display: block;\n",
       "    margin: auto;\n",
       "}\n",
       "figure img {\n",
       "    vertical-align: top;\n",
       "    text-align: center;\n",
       "    display: block;\n",
       "    margin: auto;\n",
       "}\n",
       "figure figcaption {\n",
       "    border: 0px;\n",
       "    text-align: center;\n",
       "    font-size: 11px;\n",
       "    margin: auto;\n",
       "}\n",
       "\n",
       "</style>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/markdown": [
       "**Start at:** Monday 17 June 2024, 15:08:27  \n",
       "**Hostname:** localhost.localdomain (Linux)"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/markdown": [
       "<p style=\"text-align: center\"><img width=\"800px\" src=\"./config/svg/pyPhysChemBanner.svg\" style=\"margin-left:auto; margin-right:auto\"/></p>"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import os\n",
    "import sys\n",
    "\n",
    "cwd0 = './config/'\n",
    "sys.path.append(cwd0)\n",
    "\n",
    "import visualID_Eng as vID\n",
    "from visualID_Eng import color, fg, bg, hl\n",
    "vID.init(cwd0)\n",
    "import tools4pyPhysChem as t4pPC"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ff349db3-abff-4dc8-a639-e953a5097961",
   "metadata": {},
   "source": [
    "# Basics of Artificial Neural Networks (ANNs) for supervised learning\n",
    "\n",
    "## Introduction\n",
    "\n",
    "### We will work on the iris dataset\n",
    "\n",
    "<div class=\"intro\">\n",
    "    \n",
    "<img src=\"./DS4B-Slides/perceptron3.jpg\" alt=\"perceptron3\" width=\"1200\" style=\"display: block; margin: 0 auto; align:center\"/>\n",
    "</div>\n",
    "\n",
    "### Goal\n",
    "\n",
    "<div class=\"intro\">\n",
    "\n",
    "The linear relationship between the length and width of petals has already been underlined, whatever the species\n",
    "\n",
    "<p style=\"text-align: center\"><img width=\"500px\" src=\"DS4B-svg/jointplot_petals.png\" style=\"margin-left:auto; margin-right:auto\" id=\"img_jpc\"/></p>\n",
    "\n",
    "</div>\n",
    "<br>\n",
    "<div class=\"intro\">\n",
    "\n",
    "We are going to illustrate the difference between the 1D linear regression with the least-squared techniques available in SciPy or scikit-learn and the 1D linear regression done with a single neuron, the so-called **perceptron model**.\n",
    "\n",
    "<img src=\"./DS4B-svg/SingleNeuron-dallE.png\" alt=\"perceptron\" width=\"300\" style=\"display: block; margin: 0 auto; align:center\"/>\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0c77db77-0b98-45c9-b793-07071cec6144",
   "metadata": {},
   "source": [
    "### How to do that with python? With tensorflow.keras\n",
    "\n",
    "#### Introduction\n",
    "\n",
    "<div class=\"intro\">\n",
    "    \n",
    "[TensorFlow](https://www.tensorflow.org/) is the premier open-source deep learning framework developed and maintained by Google. Although using TensorFlow directly can be challenging, the modern tf.keras Application Programming Interface (API) brings [Keras](https://keras.io/)’s simplicity and ease of use to the TensorFlow project. Keras is an open-source deep learning library written in Python. The project was started in 2015 by Francois Chollet. It quickly became a popular framework for developers, becoming one of, if not the most, popular deep learning libraries.\n",
    "\n",
    "In 2019, Google released a new version of their TensorFlow deep learning library (TensorFlow 2) that integrated the Keras API directly and promoted this interface as the default or standard interface for deep learning development on the platform.\n",
    "\n",
    "**Using tf.keras allows you to design, fit, evaluate, and use deep learning models to make predictions in just a few lines of code.** It is one reason that explains its popularity. It makes common deep learning tasks, such as classification and regression predictive modeling, accessible to average developers looking to get things done.\n",
    "\n",
    "In this tutorial, you will discover a step-by-step guide to developing a **perceptron model** in TensorFlow using the tf.keras API\n",
    "\n",
    "<img src=\"./DS4B-Slides/tf.keras.png\" alt=\"tf.keras\" width=\"500\" style=\"display: block; margin: 0 auto; align:center\"/>\n",
    "\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5f49d454-dcd5-47c5-b6df-bb683377f36f",
   "metadata": {},
   "source": [
    "#### What is the necessary mathematical prerequisite?\n",
    "\n",
    "<div class=\"intro\">\n",
    "\n",
    "Not that much, in the first place. **You do not need to know the math first**. Yet, a fair knowledge in the tools from linear algebra, probability, and statistics is necessary. From a practical point of view, you will use codes and explore algorithmic behavior with different inputs and outputs. Knowing the math will not tell you what algorithm to choose or how to configure it best. You can only discover that through careful, controlled experiments.\n",
    "\n",
    "**You do not need to know how the algorithms work**. Knowing the limitations and how to configure deep learning algorithms is important. But **learning about algorithms can come later**. You need to build up this algorithm knowledge slowly over a long period of time\n",
    "\n",
    "**You do not need to be a deep learning expert**. You can learn about the benefits and limitations of various algorithms later. There are plenty of tutorials and examples on the web that you can work on later.\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b5d9f71a-b50e-42be-b8fd-e06f32890526",
   "metadata": {},
   "source": [
    "## Linear Regression\n",
    "\n",
    "<div class=\"rq\">\n",
    "\n",
    "Before playing with the perceptron model, let's do an \"ordinary\" linear regression, which involves minimizing the sum of squared differences (least squares) between the observed and predicted value.\n",
    "</div>\n",
    "\n",
    "### \"Ordinary\" least squared fit linear regression with scikit learn\n",
    "\n",
    "#### Introduction\n",
    "\n",
    "<div class=\"intro\">\n",
    "\n",
    "`LinearRegression` is a class in the scikit-learn library, which is a popular machine learning library in Python. This class provides an implementation of linear regression for both simple and multiple linear regression problems. It uses the least squares method to estimate the coefficients. \n",
    "    \n",
    "\n",
    "<img src=\"./DS4B-Slides/perceptron4.jpg\" alt=\"perceptron4\" width=\"1200\" style=\"display: block; margin: 0 auto; align:center\"/>\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b1afe85a-b43e-43f2-b964-2a57624cff42",
   "metadata": {},
   "source": [
    "#### Workflow\n",
    "\n",
    "<div class=\"intro\">\n",
    "\n",
    "Here's a brief overview of how you can use the LinearRegression class in scikit-learn:\n",
    "\n",
    "</div>\n",
    "<br>\n",
    "<div class=\"introT\" title=\"Import the LinearRegression class and create an instance of the LinearRegression class\">\n",
    "\n",
    "<br>\n",
    "\n",
    "```python\n",
    "from sklearn.linear_model import LinearRegression\n",
    "model = LinearRegression()\n",
    "```\n",
    "</div>\n",
    "<br>\n",
    "<div class=\"introT\" title=\"Fit the model to your data\">\n",
    "\n",
    "<br>\n",
    "\n",
    "```python\n",
    "model.fit(X, y)\n",
    "```\n",
    "<br>\n",
    "X is the feature matrix (in the case of simple linear regression, it would be a one-dimensional array or column vector).\n",
    "y is the target variable.\n",
    "\n",
    "</div>\n",
    "<br>\n",
    "<div class=\"introT\" title=\"Make predictions\">\n",
    "\n",
    "<br>\n",
    "\n",
    "```python\n",
    "predictions = model.predict(X_new)\n",
    "```\n",
    "<br>\n",
    "X_new is the new data for which you want to make predictions.\n",
    "\n",
    "</div>\n",
    "<br>\n",
    "<div class=\"introT\" title=\"Access the model parameters\">\n",
    "\n",
    "<br>\n",
    "\n",
    "```python\n",
    "coefficient = model.coef_\n",
    "intercept = model.intercept_\n",
    "```\n",
    "<br>\n",
    "coef_ gives you the coefficients (slopes) of the linear regression model.\n",
    "intercept_ gives you the intercept of the model\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b6ddca59-8d64-42f0-8b5b-8b103178478e",
   "metadata": {},
   "source": [
    "#### Application\n",
    "\n",
    "<div class=\"intro\">\n",
    "    \n",
    "Linear regression between petal width and length\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "bef8c712-dd11-45b9-b0a1-4db7d69f04f7",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>sepal_length</th>\n",
       "      <th>sepal_width</th>\n",
       "      <th>petal_length</th>\n",
       "      <th>petal_width</th>\n",
       "      <th>species</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5.1</td>\n",
       "      <td>3.5</td>\n",
       "      <td>1.4</td>\n",
       "      <td>0.2</td>\n",
       "      <td>setosa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4.9</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.4</td>\n",
       "      <td>0.2</td>\n",
       "      <td>setosa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4.7</td>\n",
       "      <td>3.2</td>\n",
       "      <td>1.3</td>\n",
       "      <td>0.2</td>\n",
       "      <td>setosa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.6</td>\n",
       "      <td>3.1</td>\n",
       "      <td>1.5</td>\n",
       "      <td>0.2</td>\n",
       "      <td>setosa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5.0</td>\n",
       "      <td>3.6</td>\n",
       "      <td>1.4</td>\n",
       "      <td>0.2</td>\n",
       "      <td>setosa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>145</th>\n",
       "      <td>6.7</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.2</td>\n",
       "      <td>2.3</td>\n",
       "      <td>virginica</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>146</th>\n",
       "      <td>6.3</td>\n",
       "      <td>2.5</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.9</td>\n",
       "      <td>virginica</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>147</th>\n",
       "      <td>6.5</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>virginica</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>148</th>\n",
       "      <td>6.2</td>\n",
       "      <td>3.4</td>\n",
       "      <td>5.4</td>\n",
       "      <td>2.3</td>\n",
       "      <td>virginica</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149</th>\n",
       "      <td>5.9</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.1</td>\n",
       "      <td>1.8</td>\n",
       "      <td>virginica</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>150 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     sepal_length  sepal_width  petal_length  petal_width    species\n",
       "0             5.1          3.5           1.4          0.2     setosa\n",
       "1             4.9          3.0           1.4          0.2     setosa\n",
       "2             4.7          3.2           1.3          0.2     setosa\n",
       "3             4.6          3.1           1.5          0.2     setosa\n",
       "4             5.0          3.6           1.4          0.2     setosa\n",
       "..            ...          ...           ...          ...        ...\n",
       "145           6.7          3.0           5.2          2.3  virginica\n",
       "146           6.3          2.5           5.0          1.9  virginica\n",
       "147           6.5          3.0           5.2          2.0  virginica\n",
       "148           6.2          3.4           5.4          2.3  virginica\n",
       "149           5.9          3.0           5.1          1.8  virginica\n",
       "\n",
       "[150 rows x 5 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "a   =   2.226\n",
      "b   =   1.091\n",
      "\n",
      "r^2 = 0.92690\n",
      "MSE =   0.226\n",
      "MAE =   0.368\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "dfi=pd.read_csv('./DS4B-iris-data/iris.csv', sep=\";\")\n",
    "display(dfi)\n",
    "\n",
    "from sklearn.linear_model import LinearRegression\n",
    "pw = dfi[\"petal_width\"].values.reshape(-1,1)\n",
    "# print(pw.shape)\n",
    "# print(pw)\n",
    "pl = dfi[\"petal_length\"].values\n",
    "# creation of the LinearRegression object\n",
    "fitlin = LinearRegression()\n",
    "\n",
    "# search for the parameters a and b of the linear model\n",
    "fitlin.fit(pw, pl)\n",
    "pl_hat=fitlin.predict(pw)\n",
    "a = fitlin.coef_\n",
    "b = fitlin.intercept_\n",
    "from sklearn.metrics import r2_score\n",
    "r2 = r2_score(pl, pl_hat)\n",
    "print(\"a   ={:8.3f}\\nb   ={:8.3f}\\n\\nr^2 ={:8.5f}\".format(a[0], b, r2))\n",
    "from sklearn.metrics import mean_squared_error as MSE\n",
    "from sklearn.metrics import mean_absolute_error as MAE\n",
    "print(\"MSE ={:8.3f}\\nMAE ={:8.3f}\".format(MSE(pl,pl_hat), MAE(pl,pl_hat)))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "df8a32d1-551f-4851-94fe-eb7a28c5d6bc",
   "metadata": {},
   "source": [
    "<div class=\"rq\">\n",
    "\n",
    "The formula $\\hat{y}=2.226x+1.091$ is the law that allows to predict the length of a petal as a functions of its width\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "99239dcc-1c35-4356-bf0f-f3dc69f13c66",
   "metadata": {},
   "source": [
    "<div class=\"rqE\">\n",
    "\n",
    "The bad news is that the analytical formula\n",
    "\n",
    "$${\\hat{\\boldsymbol{\\theta}}=(\\textbf{X}^{\\mathrm{T}}.\\textbf{X})^{-1}.\\textbf{X}^{\\mathrm{T}}.\\textbf{Y}}$$\n",
    "\n",
    "presents a high complexity and cost for real multivariate function $f(x_{1},x_{2},x_{3},...)$\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c4027ae9-4cbc-4195-bca7-4ea9661c4d3b",
   "metadata": {},
   "source": [
    "### Linear regresion with a single neuron: the perceptron model\n",
    "\n",
    "#### The seminal perceptron model\n",
    "\n",
    "<div class=\"intro\">\n",
    "\n",
    "It is shown below. It is a simple binary classification algorithm and serves as the fundamental building block of neural networks. Developed by Frank Rosenblatt in 1957, a perceptron takes a set of binary input features and produces a binary output, making it a binary linear classifier. The model consists of input nodes (features), weights associated with each input, a summation function, an activation function (typically a step function), and an output.\n",
    "\n",
    "The perceptron's learning is based on the adjustment of weights during training to minimize classification errors. While a single perceptron has limitations, it becomes more powerful when combined into multilayer perceptrons, forming the foundation for more sophisticated neural network architectures. The perceptron laid the groundwork for the development of artificial neural networks and the broader field of deep learning.\n",
    "    \n",
    "<img src=\"./DS4B-Slides/perceptron6.jpg\" alt=\"perceptron6\" width=\"1200\" style=\"display: block; margin: 0 auto; align:center\"/>\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c822c799-0d0b-486f-b3fe-7d9b8c300770",
   "metadata": {},
   "source": [
    "#### A perceptron with a single input value and no action (activation) potential\n",
    "\n",
    "<div class=\"intro\">\n",
    "\n",
    "A single neuron is enough for predicting the length of a petal as a function of its width, given that there seems to be a linear relationship between bowth properties. It's similar to the <a href=\"https://en.wikipedia.org/wiki/Perceptron\">perceptron invented by Rosenblatt</a> in 1958, except that here we're going to optimise weight and bias with the backpropagation algorithm, which is more recent (1974) and whose principle has renewed interest in neural networks. The perceptron is a single layer feed-forward neural network.\n",
    "\n",
    "This will also give us the opportunity to focus on the quite simple mathematical modelling of a neuron, summarised in the figure below \n",
    "<br><br>\n",
    "<p style=\"text-align: center\"><img width=\"350px\" src=\"DS4B-svg/1neuron-perceptron-wBackP.png\" style=\"margin-left:auto; margin-right:auto\" id=\"img_perceptron\"/></p>\n",
    "\n",
    "The neuron receives the input information *x* and optimising the weight of the neuronal connection (*w*) between it and the input neuron will enable it to reproduce the value to be reproduced (*y*) as well as possible, with the bias (*b*) allowing this fit to be adjusted. This optimisation scheme is carried out for all the training data. \n",
    "\n",
    "**Backpropagation**, short for \"backward propagation of errors,\" is a supervised learning algorithm used in the **training** of this neuron. It is a key optimization algorithm for adjusting the neuron's weight and bias to minimize the difference between the predicted output and the actual output. Using a knowledge acquisition vocabulary, it gives a **feedback** to the model.\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "87da9312-07eb-45c6-95b6-8b3999033fc5",
   "metadata": {},
   "source": [
    "#### Workflow\n",
    "\n",
    "<div class=\"intro\">\n",
    "\n",
    "It is summarized in the figure below\n",
    "\n",
    "<p style=\"text-align: center\"><img width=\"700px\" src=\"DS4B-svg/BilanML-Classification-English.png\" style=\"margin-left:auto; margin-right:auto\" id=\"workflow\"/></p>\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f25610cc-87e2-43e3-84a9-f4520efde667",
   "metadata": {},
   "source": [
    "#### Data reading and importation of libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "be1740be-d21c-4254-95ee-ac5f5da2d6bb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dfi. Structure (shape) :(150, 5)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>sepal_length</th>\n",
       "      <th>sepal_width</th>\n",
       "      <th>petal_length</th>\n",
       "      <th>petal_width</th>\n",
       "      <th>species</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5.1</td>\n",
       "      <td>3.5</td>\n",
       "      <td>1.4</td>\n",
       "      <td>0.2</td>\n",
       "      <td>setosa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4.9</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.4</td>\n",
       "      <td>0.2</td>\n",
       "      <td>setosa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4.7</td>\n",
       "      <td>3.2</td>\n",
       "      <td>1.3</td>\n",
       "      <td>0.2</td>\n",
       "      <td>setosa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.6</td>\n",
       "      <td>3.1</td>\n",
       "      <td>1.5</td>\n",
       "      <td>0.2</td>\n",
       "      <td>setosa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5.0</td>\n",
       "      <td>3.6</td>\n",
       "      <td>1.4</td>\n",
       "      <td>0.2</td>\n",
       "      <td>setosa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>145</th>\n",
       "      <td>6.7</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.2</td>\n",
       "      <td>2.3</td>\n",
       "      <td>virginica</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>146</th>\n",
       "      <td>6.3</td>\n",
       "      <td>2.5</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.9</td>\n",
       "      <td>virginica</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>147</th>\n",
       "      <td>6.5</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>virginica</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>148</th>\n",
       "      <td>6.2</td>\n",
       "      <td>3.4</td>\n",
       "      <td>5.4</td>\n",
       "      <td>2.3</td>\n",
       "      <td>virginica</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149</th>\n",
       "      <td>5.9</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.1</td>\n",
       "      <td>1.8</td>\n",
       "      <td>virginica</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>150 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     sepal_length  sepal_width  petal_length  petal_width    species\n",
       "0             5.1          3.5           1.4          0.2     setosa\n",
       "1             4.9          3.0           1.4          0.2     setosa\n",
       "2             4.7          3.2           1.3          0.2     setosa\n",
       "3             4.6          3.1           1.5          0.2     setosa\n",
       "4             5.0          3.6           1.4          0.2     setosa\n",
       "..            ...          ...           ...          ...        ...\n",
       "145           6.7          3.0           5.2          2.3  virginica\n",
       "146           6.3          2.5           5.0          1.9  virginica\n",
       "147           6.5          3.0           5.2          2.0  virginica\n",
       "148           6.2          3.4           5.4          2.3  virginica\n",
       "149           5.9          3.0           5.1          1.8  virginica\n",
       "\n",
       "[150 rows x 5 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np \n",
    "import seaborn as sns\n",
    "import pandas as pd\n",
    "import random\n",
    "import matplotlib.pyplot as plt\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "from IPython.display import display\n",
    "\n",
    "\n",
    "dfi=pd.read_csv('./DS4B-iris-data/iris.csv', sep=\";\")\n",
    "print(f\"Dfi. Structure (shape) :{dfi.shape}\")\n",
    "display(dfi)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2a902bd8-96e8-4a21-9bb7-74fbde62e973",
   "metadata": {},
   "source": [
    "#### Holdout method\n",
    "\n",
    "<div class=\"intro\">\n",
    "\n",
    "Model evaluation using the hold-out method entails **splitting the dataset** into **training** and **test** datasets in order to evaluate the performance of the model on data it has not been trained on\n",
    "\n",
    "</div>\n",
    "<br>\n",
    "<div class=\"rq\">\n",
    "\n",
    "**It's exactly like a human being's learning process**. Students learns a certain number of concepts, equations, ideas and applications. They are then assessed on their ability to deal correctly with a new problem within the field of application of what they have learned. \n",
    "\n",
    "A \"simple\" machine learning algorithm does not learn concepts. It learns from a large number of examples to **make the link** between input properties *X*, called **descriptors or features**, and one or more related properties *y*, also called **attributes**. \n",
    "\n",
    "How this is linked is generally obscure, although there are algorithms that can help identify which descriptors the properties to be reproduced seem to depend on the most.\n",
    "</div>\n",
    "<br>\n",
    "<div class=\"intro\">\n",
    "\n",
    "<img src=\"./DS4B-svg/Basic-holdout-English.svg\" alt=\"holdout\" width=\"400\" style=\"display: block; margin: 0 auto; align:center\"/>\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "0302933c-65c3-43d4-ba01-93ccf0adceca",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_train = dfi.sample(frac=0.8, axis='index') \n",
    "data_test  = dfi.drop(data_train.index)\n",
    "x_train = data_train[['petal_width']]\n",
    "y_train = data_train[['petal_length']]\n",
    "x_test = data_test[['petal_width']]\n",
    "y_test = data_test[['petal_length']]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "895a0b5c-1502-4dfa-8b0b-fe413c180650",
   "metadata": {},
   "source": [
    "#### Definition and training of the model\n",
    "\n",
    "<div class=\"intro\">\n",
    "\n",
    "The ML model development actually requires to:\n",
    "- Define the model\n",
    "- Compile the model\n",
    "- Fit the model\n",
    "- Evaluate the model\n",
    "- Make predictions\n",
    "  \n",
    "</div>\n",
    "<br>\n",
    "<div class=\"introT\" title=\"Define the model\">\n",
    "\n",
    "We will use the **sequential model API**, that is the simplest and the recommended API, especially when getting started. It is referred to as “sequential” because it involves defining a [Sequential class](https://keras.io/api/models/sequential/) and adding layers to the model one by one in a linear manner, from input to output. The perceptron example below defines a Sequential MLP model that accepts one input and that has an output layer with one node to predict a numerical value. \n",
    "\n",
    "\n",
    "```python\n",
    "# define the model\n",
    "model = keras.models.Sequential()\n",
    "model.add(keras.layers.Input(1, name='inpLayer'))\n",
    "model.add(keras.layers.Dense(1, name='outLayer'))\n",
    "```\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f9a97dce-1159-44a5-ac70-41763cc2d552",
   "metadata": {},
   "source": [
    "<div class=\"introT\" title=\"Compile the model\">\n",
    "\n",
    "Optimizing the model consists in finding the optimal weights and biases that fit the expected outcome of the ANN:\n",
    "<img src=\"./DS4B-Slides/perceptron7.jpg\" alt=\"perceptron7\" width=\"1200\" style=\"display: block; margin: 0 auto; align:center\"/>\n",
    "\n",
    "\n",
    "Prior to the optimization, it is necessary to compile the model. It requires first selecting the loss function that you want to optimize, such as mean squared error (MSE). It also requires that you select an algorithm to perform the optimization procedure, typically stochastic gradient descent or a modern variation, such as Adam.\n",
    "\n",
    "```python\n",
    "# compile the model\n",
    "model.compile(optimizer = 'adam',\n",
    "              loss      = 'mse',\n",
    "              metrics   = ['mae', 'mse'] )\n",
    "\n",
    "```\n",
    "</div>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "10abe212-ddfe-4bb0-966b-01d85d01616d",
   "metadata": {},
   "source": [
    "<div class=\"introT\" title=\"Fit the model\">\n",
    "    \n",
    "Fitting the model requires that you first select the training configuration, such as:\n",
    "- the number of epochs = loops through the training dataset = iterations of the optimizer, as shown in the figure above\n",
    "- the batch size = number of samples in an epoch used to estimate model error\n",
    "\n",
    "**Training** is another term that designates this fitting part. The chosen optimization algorithm is actually applied to minimize the chosen loss function and updates the model using the backpropagation of the error algorithm.\n",
    "\n",
    "Fitting the model is the slow part of the whole process and can take seconds to hours to days, depending on the complexity of the model, the hardware you’re using, and the size of the training dataset\n",
    "\n",
    "``` python\n",
    "# fit the model\n",
    "ANNhistory = ANNmodel.fit(x_train,\n",
    "                    y_train,\n",
    "                    epochs          = 250,\n",
    "                    batch_size      = 10,\n",
    "                    verbose         = 1,\n",
    "                    validation_data = (x_test, y_test),\n",
    "                    callbacks=[es])\n",
    "```\n",
    "\n",
    "A callback was added, namely\n",
    "\n",
    "```python\n",
    "es = EarlyStopping(monitor='val_loss', mode='min', verbose=1, patience=10)\n",
    "```\n",
    "\n",
    "<br>\n",
    "The `EarlyStopping` callback stops training when a monitored metric has stopped improving. It's nothing else than a threshold to stop the optimization process.\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "adc63212-de34-4876-9528-b637967d31fd",
   "metadata": {},
   "source": [
    "<div class=\"introT\" title=\"Evaluate the model\">\n",
    "\n",
    "Evaluating the model requires that you first choose a **holdout dataset** used to evaluate the model. This should be data not used in the training process so that you can get an unbiased estimate of the performance of the model when making predictions on new data.\n",
    "\n",
    "The speed of a model evaluation is proportional to the amount of data you want to use for the evaluation, although it is much faster than training as the model is not changed.\n",
    "\n",
    "From an API perspective, this involves calling a function with the holdout dataset and getting a loss and perhaps other metrics that can be reported.\n",
    "\n",
    "The **accuracy of the model** will be demonstrated in the next subsection. \n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c75b2496-d860-4418-9989-954e85166b95",
   "metadata": {},
   "source": [
    "<div class=\"introT\" title=\"Make predictions\">\n",
    "\n",
    "Making a prediction is the final step in the life cycle. It is **why we wanted the model in the first place**.\n",
    "\n",
    "It requires you have new data for which a prediction is required, e.g., where you do not have the target values.\n",
    "\n",
    "From an API perspective, you simply call a function to make a prediction of a class label, probability, or numerical value—whatever you designed your model to predict.\n",
    "\n",
    "```python\n",
    "ANNmodel.predict(x)\n",
    "```\n",
    "</div>\n",
    "<br>\n",
    "<div class=\"rq\">\n",
    "\n",
    "Let's start with the definition of the model and its fit\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "322c466c-1206-463f-96e4-3094ed9c630d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"sequential_1\"</span>\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1mModel: \"sequential_1\"\u001b[0m\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
       "┃<span style=\"font-weight: bold\"> Layer (type)                    </span>┃<span style=\"font-weight: bold\"> Output Shape           </span>┃<span style=\"font-weight: bold\">       Param # </span>┃\n",
       "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
       "│ outLayer (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">1</span>)              │             <span style=\"color: #00af00; text-decoration-color: #00af00\">2</span> │\n",
       "└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
       "</pre>\n"
      ],
      "text/plain": [
       "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
       "┃\u001b[1m \u001b[0m\u001b[1mLayer (type)                   \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape          \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m      Param #\u001b[0m\u001b[1m \u001b[0m┃\n",
       "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
       "│ outLayer (\u001b[38;5;33mDense\u001b[0m)                │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1\u001b[0m)              │             \u001b[38;5;34m2\u001b[0m │\n",
       "└─────────────────────────────────┴────────────────────────┴───────────────┘\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">2</span> (8.00 B)\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1m Total params: \u001b[0m\u001b[38;5;34m2\u001b[0m (8.00 B)\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">2</span> (8.00 B)\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m2\u001b[0m (8.00 B)\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> (0.00 B)\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def define_1ANN():\n",
    "    \n",
    "    model = keras.models.Sequential()\n",
    "    model.add(keras.layers.Input(shape=(1,), name='inpLayer'))\n",
    "    model.add(keras.layers.Dense(1, name='outLayer'))\n",
    "    \n",
    "    model.compile(optimizer = 'adam',\n",
    "                  loss      = 'mse',\n",
    "                  metrics   = ['mae', 'mse'] )\n",
    "    return model\n",
    "\n",
    "from tensorflow.keras.callbacks import EarlyStopping\n",
    "es = EarlyStopping(monitor='val_loss', mode='min', verbose=1, patience=10)\n",
    "ANNmodel=define_1ANN()\n",
    "ANNmodel.summary()\n",
    "\n",
    "from tensorflow.keras.utils import plot_model\n",
    "plot_model(ANNmodel, show_shapes=True, show_layer_names=True, show_layer_activations=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "15ae89da-4fac-4de2-bdea-a5e0f9bafd3b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 12ms/step - loss: 6.1206 - mae: 2.2863 - mse: 6.1206 - val_loss: 6.5352 - val_mae: 2.4074 - val_mse: 6.5352\n",
      "Epoch 2/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 5.8011 - mae: 2.2266 - mse: 5.8011 - val_loss: 6.3926 - val_mae: 2.3802 - val_mse: 6.3926\n",
      "Epoch 3/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 6.1503 - mae: 2.3023 - mse: 6.1503 - val_loss: 6.2510 - val_mae: 2.3529 - val_mse: 6.2510\n",
      "Epoch 4/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 5.4419 - mae: 2.1607 - mse: 5.4419 - val_loss: 6.1121 - val_mae: 2.3259 - val_mse: 6.1121\n",
      "Epoch 5/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 5.2698 - mae: 2.1301 - mse: 5.2698 - val_loss: 5.9767 - val_mae: 2.2992 - val_mse: 5.9767\n",
      "Epoch 6/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 5.7801 - mae: 2.2376 - mse: 5.7801 - val_loss: 5.8405 - val_mae: 2.2720 - val_mse: 5.8405\n",
      "Epoch 7/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 5.1516 - mae: 2.1077 - mse: 5.1516 - val_loss: 5.7094 - val_mae: 2.2456 - val_mse: 5.7094\n",
      "Epoch 8/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 5.7017 - mae: 2.2293 - mse: 5.7017 - val_loss: 5.5800 - val_mae: 2.2191 - val_mse: 5.5800\n",
      "Epoch 9/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 5.0368 - mae: 2.0814 - mse: 5.0368 - val_loss: 5.4527 - val_mae: 2.1928 - val_mse: 5.4527\n",
      "Epoch 10/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 5.1268 - mae: 2.1031 - mse: 5.1268 - val_loss: 5.3302 - val_mae: 2.1672 - val_mse: 5.3302\n",
      "Epoch 11/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 4.7469 - mae: 2.0048 - mse: 4.7469 - val_loss: 5.2082 - val_mae: 2.1413 - val_mse: 5.2082\n",
      "Epoch 12/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 4.4761 - mae: 1.9586 - mse: 4.4761 - val_loss: 5.0896 - val_mae: 2.1159 - val_mse: 5.0896\n",
      "Epoch 13/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 4.7077 - mae: 2.0149 - mse: 4.7077 - val_loss: 4.9697 - val_mae: 2.0898 - val_mse: 4.9697\n",
      "Epoch 14/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 3.9789 - mae: 1.8230 - mse: 3.9789 - val_loss: 4.8559 - val_mae: 2.0648 - val_mse: 4.8559\n",
      "Epoch 15/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 4.4993 - mae: 1.9581 - mse: 4.4993 - val_loss: 4.7402 - val_mae: 2.0391 - val_mse: 4.7402\n",
      "Epoch 16/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 4.1372 - mae: 1.8579 - mse: 4.1372 - val_loss: 4.6289 - val_mae: 2.0140 - val_mse: 4.6289\n",
      "Epoch 17/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 4.2118 - mae: 1.9068 - mse: 4.2118 - val_loss: 4.5203 - val_mae: 1.9892 - val_mse: 4.5203\n",
      "Epoch 18/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 4.1040 - mae: 1.8616 - mse: 4.1040 - val_loss: 4.4138 - val_mae: 1.9645 - val_mse: 4.4138\n",
      "Epoch 19/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 3.5980 - mae: 1.7291 - mse: 3.5980 - val_loss: 4.3102 - val_mae: 1.9403 - val_mse: 4.3102\n",
      "Epoch 20/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 4.1778 - mae: 1.8739 - mse: 4.1778 - val_loss: 4.2045 - val_mae: 1.9152 - val_mse: 4.2045\n",
      "Epoch 21/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 3.6218 - mae: 1.7484 - mse: 3.6218 - val_loss: 4.1044 - val_mae: 1.8911 - val_mse: 4.1044\n",
      "Epoch 22/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 3.6192 - mae: 1.7367 - mse: 3.6192 - val_loss: 4.0051 - val_mae: 1.8669 - val_mse: 4.0051\n",
      "Epoch 23/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 3.5993 - mae: 1.7337 - mse: 3.5993 - val_loss: 3.9072 - val_mae: 1.8428 - val_mse: 3.9072\n",
      "Epoch 24/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 3.5147 - mae: 1.7086 - mse: 3.5147 - val_loss: 3.8119 - val_mae: 1.8189 - val_mse: 3.8119\n",
      "Epoch 25/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 3.3488 - mae: 1.6586 - mse: 3.3488 - val_loss: 3.7192 - val_mae: 1.7954 - val_mse: 3.7192\n",
      "Epoch 26/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 3.3005 - mae: 1.6706 - mse: 3.3005 - val_loss: 3.6271 - val_mae: 1.7718 - val_mse: 3.6271\n",
      "Epoch 27/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 2.8374 - mae: 1.5314 - mse: 2.8374 - val_loss: 3.5385 - val_mae: 1.7487 - val_mse: 3.5385\n",
      "Epoch 28/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 3.3799 - mae: 1.6998 - mse: 3.3799 - val_loss: 3.4493 - val_mae: 1.7252 - val_mse: 3.4493\n",
      "Epoch 29/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 3.3783 - mae: 1.6771 - mse: 3.3783 - val_loss: 3.3616 - val_mae: 1.7018 - val_mse: 3.3616\n",
      "Epoch 30/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 3.2269 - mae: 1.6462 - mse: 3.2269 - val_loss: 3.2776 - val_mae: 1.6790 - val_mse: 3.2776\n",
      "Epoch 31/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 2.9447 - mae: 1.5659 - mse: 2.9447 - val_loss: 3.1962 - val_mae: 1.6566 - val_mse: 3.1962\n",
      "Epoch 32/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 2.6594 - mae: 1.4590 - mse: 2.6594 - val_loss: 3.1166 - val_mae: 1.6343 - val_mse: 3.1166\n",
      "Epoch 33/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 2.8029 - mae: 1.5391 - mse: 2.8029 - val_loss: 3.0364 - val_mae: 1.6117 - val_mse: 3.0364\n",
      "Epoch 34/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 2.6881 - mae: 1.4910 - mse: 2.6881 - val_loss: 2.9595 - val_mae: 1.5896 - val_mse: 2.9595\n",
      "Epoch 35/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 2.8635 - mae: 1.5309 - mse: 2.8635 - val_loss: 2.8827 - val_mae: 1.5673 - val_mse: 2.8827\n",
      "Epoch 36/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 2.5263 - mae: 1.4263 - mse: 2.5263 - val_loss: 2.8082 - val_mae: 1.5453 - val_mse: 2.8082\n",
      "Epoch 37/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 2.5902 - mae: 1.4546 - mse: 2.5902 - val_loss: 2.7358 - val_mae: 1.5236 - val_mse: 2.7358\n",
      "Epoch 38/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 2.5150 - mae: 1.4182 - mse: 2.5150 - val_loss: 2.6649 - val_mae: 1.5020 - val_mse: 2.6649\n",
      "Epoch 39/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 2.6442 - mae: 1.4872 - mse: 2.6442 - val_loss: 2.5940 - val_mae: 1.4801 - val_mse: 2.5940\n",
      "Epoch 40/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 2.4645 - mae: 1.4230 - mse: 2.4645 - val_loss: 2.5268 - val_mae: 1.4591 - val_mse: 2.5268\n",
      "Epoch 41/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 2.5543 - mae: 1.4583 - mse: 2.5543 - val_loss: 2.4583 - val_mae: 1.4373 - val_mse: 2.4583\n",
      "Epoch 42/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 2.0739 - mae: 1.2930 - mse: 2.0739 - val_loss: 2.3956 - val_mae: 1.4171 - val_mse: 2.3956\n",
      "Epoch 43/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 2.2578 - mae: 1.3572 - mse: 2.2578 - val_loss: 2.3324 - val_mae: 1.3964 - val_mse: 2.3324\n",
      "Epoch 44/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 2.0288 - mae: 1.2637 - mse: 2.0288 - val_loss: 2.2709 - val_mae: 1.3759 - val_mse: 2.2709\n",
      "Epoch 45/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 2.2363 - mae: 1.3251 - mse: 2.2363 - val_loss: 2.2097 - val_mae: 1.3553 - val_mse: 2.2097\n",
      "Epoch 46/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 2.1475 - mae: 1.3064 - mse: 2.1475 - val_loss: 2.1503 - val_mae: 1.3349 - val_mse: 2.1503\n",
      "Epoch 47/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.8898 - mae: 1.2181 - mse: 1.8898 - val_loss: 2.0937 - val_mae: 1.3152 - val_mse: 2.0937\n",
      "Epoch 48/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 2.0149 - mae: 1.2784 - mse: 2.0149 - val_loss: 2.0365 - val_mae: 1.2950 - val_mse: 2.0365\n",
      "Epoch 49/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.8139 - mae: 1.1998 - mse: 1.8139 - val_loss: 1.9815 - val_mae: 1.2752 - val_mse: 1.9815\n",
      "Epoch 50/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 2.2094 - mae: 1.3436 - mse: 2.2094 - val_loss: 1.9266 - val_mae: 1.2551 - val_mse: 1.9266\n",
      "Epoch 51/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.7971 - mae: 1.1818 - mse: 1.7971 - val_loss: 1.8759 - val_mae: 1.2363 - val_mse: 1.8759\n",
      "Epoch 52/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.8015 - mae: 1.2013 - mse: 1.8015 - val_loss: 1.8241 - val_mae: 1.2168 - val_mse: 1.8241\n",
      "Epoch 53/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.5715 - mae: 1.0945 - mse: 1.5715 - val_loss: 1.7753 - val_mae: 1.1981 - val_mse: 1.7753\n",
      "Epoch 54/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.8815 - mae: 1.2220 - mse: 1.8815 - val_loss: 1.7248 - val_mae: 1.1784 - val_mse: 1.7248\n",
      "Epoch 55/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.5497 - mae: 1.1012 - mse: 1.5497 - val_loss: 1.6783 - val_mae: 1.1600 - val_mse: 1.6783\n",
      "Epoch 56/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.5438 - mae: 1.0959 - mse: 1.5438 - val_loss: 1.6325 - val_mae: 1.1415 - val_mse: 1.6325\n",
      "Epoch 57/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.4884 - mae: 1.0533 - mse: 1.4884 - val_loss: 1.5879 - val_mae: 1.1233 - val_mse: 1.5879\n",
      "Epoch 58/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.2462 - mae: 0.9608 - mse: 1.2462 - val_loss: 1.5450 - val_mae: 1.1054 - val_mse: 1.5450\n",
      "Epoch 59/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.4913 - mae: 1.0402 - mse: 1.4913 - val_loss: 1.5013 - val_mae: 1.0869 - val_mse: 1.5013\n",
      "Epoch 60/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.2452 - mae: 0.9399 - mse: 1.2452 - val_loss: 1.4601 - val_mae: 1.0691 - val_mse: 1.4601\n",
      "Epoch 61/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.4127 - mae: 1.0050 - mse: 1.4127 - val_loss: 1.4195 - val_mae: 1.0513 - val_mse: 1.4195\n",
      "Epoch 62/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.2530 - mae: 0.9818 - mse: 1.2530 - val_loss: 1.3795 - val_mae: 1.0335 - val_mse: 1.3795\n",
      "Epoch 63/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.2707 - mae: 0.9655 - mse: 1.2707 - val_loss: 1.3412 - val_mae: 1.0161 - val_mse: 1.3412\n",
      "Epoch 64/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.0855 - mae: 0.8950 - mse: 1.0855 - val_loss: 1.3042 - val_mae: 0.9989 - val_mse: 1.3042\n",
      "Epoch 65/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.2988 - mae: 0.9969 - mse: 1.2988 - val_loss: 1.2668 - val_mae: 0.9814 - val_mse: 1.2668\n",
      "Epoch 66/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.1546 - mae: 0.9168 - mse: 1.1546 - val_loss: 1.2328 - val_mae: 0.9651 - val_mse: 1.2328\n",
      "Epoch 67/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.1802 - mae: 0.9226 - mse: 1.1802 - val_loss: 1.1981 - val_mae: 0.9481 - val_mse: 1.1981\n",
      "Epoch 68/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.2349 - mae: 0.9642 - mse: 1.2349 - val_loss: 1.1640 - val_mae: 0.9312 - val_mse: 1.1640\n",
      "Epoch 69/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.0643 - mae: 0.8707 - mse: 1.0643 - val_loss: 1.1323 - val_mae: 0.9151 - val_mse: 1.1323\n",
      "Epoch 70/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.8716 - mae: 0.7904 - mse: 0.8716 - val_loss: 1.1015 - val_mae: 0.8992 - val_mse: 1.1015\n",
      "Epoch 71/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.0435 - mae: 0.8797 - mse: 1.0435 - val_loss: 1.0702 - val_mae: 0.8828 - val_mse: 1.0702\n",
      "Epoch 72/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.0516 - mae: 0.8684 - mse: 1.0516 - val_loss: 1.0401 - val_mae: 0.8667 - val_mse: 1.0401\n",
      "Epoch 73/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.0930 - mae: 0.8708 - mse: 1.0930 - val_loss: 1.0112 - val_mae: 0.8509 - val_mse: 1.0112\n",
      "Epoch 74/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.8871 - mae: 0.8120 - mse: 0.8871 - val_loss: 0.9836 - val_mae: 0.8355 - val_mse: 0.9836\n",
      "Epoch 75/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.9009 - mae: 0.7931 - mse: 0.9009 - val_loss: 0.9566 - val_mae: 0.8202 - val_mse: 0.9566\n",
      "Epoch 76/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.8542 - mae: 0.7715 - mse: 0.8542 - val_loss: 0.9304 - val_mae: 0.8051 - val_mse: 0.9304\n",
      "Epoch 77/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.8935 - mae: 0.7868 - mse: 0.8935 - val_loss: 0.9047 - val_mae: 0.7899 - val_mse: 0.9047\n",
      "Epoch 78/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.8434 - mae: 0.7628 - mse: 0.8434 - val_loss: 0.8802 - val_mae: 0.7751 - val_mse: 0.8802\n",
      "Epoch 79/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.8749 - mae: 0.7815 - mse: 0.8749 - val_loss: 0.8559 - val_mae: 0.7602 - val_mse: 0.8559\n",
      "Epoch 80/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.8336 - mae: 0.7397 - mse: 0.8336 - val_loss: 0.8329 - val_mae: 0.7458 - val_mse: 0.8329\n",
      "Epoch 81/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.7586 - mae: 0.7169 - mse: 0.7586 - val_loss: 0.8106 - val_mae: 0.7315 - val_mse: 0.8106\n",
      "Epoch 82/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.6830 - mae: 0.6641 - mse: 0.6830 - val_loss: 0.7893 - val_mae: 0.7176 - val_mse: 0.7893\n",
      "Epoch 83/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.6894 - mae: 0.6865 - mse: 0.6894 - val_loss: 0.7679 - val_mae: 0.7034 - val_mse: 0.7679\n",
      "Epoch 84/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.7733 - mae: 0.7170 - mse: 0.7733 - val_loss: 0.7473 - val_mae: 0.6893 - val_mse: 0.7473\n",
      "Epoch 85/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.5890 - mae: 0.6237 - mse: 0.5890 - val_loss: 0.7281 - val_mae: 0.6760 - val_mse: 0.7281\n",
      "Epoch 86/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.6515 - mae: 0.6621 - mse: 0.6515 - val_loss: 0.7089 - val_mae: 0.6623 - val_mse: 0.7089\n",
      "Epoch 87/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.6482 - mae: 0.6451 - mse: 0.6482 - val_loss: 0.6902 - val_mae: 0.6488 - val_mse: 0.6902\n",
      "Epoch 88/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.7210 - mae: 0.7051 - mse: 0.7210 - val_loss: 0.6718 - val_mae: 0.6355 - val_mse: 0.6718\n",
      "Epoch 89/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.6077 - mae: 0.6365 - mse: 0.6077 - val_loss: 0.6553 - val_mae: 0.6241 - val_mse: 0.6553\n",
      "Epoch 90/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.5966 - mae: 0.6201 - mse: 0.5966 - val_loss: 0.6387 - val_mae: 0.6126 - val_mse: 0.6387\n",
      "Epoch 91/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 0.5398 - mae: 0.5914 - mse: 0.5398 - val_loss: 0.6228 - val_mae: 0.6012 - val_mse: 0.6228\n",
      "Epoch 92/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.6531 - mae: 0.6644 - mse: 0.6531 - val_loss: 0.6068 - val_mae: 0.5896 - val_mse: 0.6068\n",
      "Epoch 93/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.5648 - mae: 0.5975 - mse: 0.5648 - val_loss: 0.5922 - val_mae: 0.5787 - val_mse: 0.5922\n",
      "Epoch 94/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.6421 - mae: 0.6528 - mse: 0.6421 - val_loss: 0.5773 - val_mae: 0.5685 - val_mse: 0.5773\n",
      "Epoch 95/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.6523 - mae: 0.6454 - mse: 0.6523 - val_loss: 0.5639 - val_mae: 0.5591 - val_mse: 0.5639\n",
      "Epoch 96/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.6051 - mae: 0.6128 - mse: 0.6051 - val_loss: 0.5502 - val_mae: 0.5493 - val_mse: 0.5502\n",
      "Epoch 97/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.5153 - mae: 0.5818 - mse: 0.5153 - val_loss: 0.5374 - val_mae: 0.5399 - val_mse: 0.5374\n",
      "Epoch 98/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.4496 - mae: 0.5339 - mse: 0.4496 - val_loss: 0.5255 - val_mae: 0.5309 - val_mse: 0.5255\n",
      "Epoch 99/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.5055 - mae: 0.5849 - mse: 0.5055 - val_loss: 0.5132 - val_mae: 0.5215 - val_mse: 0.5132\n",
      "Epoch 100/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.4668 - mae: 0.5514 - mse: 0.4668 - val_loss: 0.5021 - val_mae: 0.5128 - val_mse: 0.5021\n",
      "Epoch 101/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.4831 - mae: 0.5680 - mse: 0.4831 - val_loss: 0.4910 - val_mae: 0.5037 - val_mse: 0.4910\n",
      "Epoch 102/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.4473 - mae: 0.5393 - mse: 0.4473 - val_loss: 0.4806 - val_mae: 0.4951 - val_mse: 0.4806\n",
      "Epoch 103/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.4382 - mae: 0.5431 - mse: 0.4382 - val_loss: 0.4703 - val_mae: 0.4864 - val_mse: 0.4703\n",
      "Epoch 104/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.4537 - mae: 0.5326 - mse: 0.4537 - val_loss: 0.4605 - val_mae: 0.4778 - val_mse: 0.4605\n",
      "Epoch 105/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.3096 - mae: 0.4369 - mse: 0.3096 - val_loss: 0.4515 - val_mae: 0.4700 - val_mse: 0.4515\n",
      "Epoch 106/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.4443 - mae: 0.5153 - mse: 0.4443 - val_loss: 0.4421 - val_mae: 0.4624 - val_mse: 0.4421\n",
      "Epoch 107/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.4421 - mae: 0.5248 - mse: 0.4421 - val_loss: 0.4334 - val_mae: 0.4557 - val_mse: 0.4334\n",
      "Epoch 108/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.4409 - mae: 0.5196 - mse: 0.4409 - val_loss: 0.4253 - val_mae: 0.4500 - val_mse: 0.4253\n",
      "Epoch 109/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.3682 - mae: 0.4797 - mse: 0.3682 - val_loss: 0.4170 - val_mae: 0.4442 - val_mse: 0.4170\n",
      "Epoch 110/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.3586 - mae: 0.4632 - mse: 0.3586 - val_loss: 0.4095 - val_mae: 0.4390 - val_mse: 0.4095\n",
      "Epoch 111/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.4090 - mae: 0.5010 - mse: 0.4090 - val_loss: 0.4021 - val_mae: 0.4337 - val_mse: 0.4021\n",
      "Epoch 112/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.3964 - mae: 0.4945 - mse: 0.3964 - val_loss: 0.3953 - val_mae: 0.4288 - val_mse: 0.3953\n",
      "Epoch 113/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.3601 - mae: 0.4672 - mse: 0.3601 - val_loss: 0.3889 - val_mae: 0.4241 - val_mse: 0.3889\n",
      "Epoch 114/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.3925 - mae: 0.5000 - mse: 0.3925 - val_loss: 0.3821 - val_mae: 0.4198 - val_mse: 0.3821\n",
      "Epoch 115/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.3505 - mae: 0.4673 - mse: 0.3505 - val_loss: 0.3764 - val_mae: 0.4161 - val_mse: 0.3764\n",
      "Epoch 116/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.3871 - mae: 0.4919 - mse: 0.3871 - val_loss: 0.3703 - val_mae: 0.4120 - val_mse: 0.3703\n",
      "Epoch 117/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.3387 - mae: 0.4563 - mse: 0.3387 - val_loss: 0.3648 - val_mae: 0.4081 - val_mse: 0.3648\n",
      "Epoch 118/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.3844 - mae: 0.4847 - mse: 0.3844 - val_loss: 0.3596 - val_mae: 0.4045 - val_mse: 0.3596\n",
      "Epoch 119/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.3098 - mae: 0.4460 - mse: 0.3098 - val_loss: 0.3546 - val_mae: 0.4015 - val_mse: 0.3546\n",
      "Epoch 120/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.3372 - mae: 0.4677 - mse: 0.3372 - val_loss: 0.3497 - val_mae: 0.3984 - val_mse: 0.3497\n",
      "Epoch 121/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2803 - mae: 0.4110 - mse: 0.2803 - val_loss: 0.3451 - val_mae: 0.3954 - val_mse: 0.3451\n",
      "Epoch 122/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.3505 - mae: 0.4510 - mse: 0.3505 - val_loss: 0.3407 - val_mae: 0.3925 - val_mse: 0.3407\n",
      "Epoch 123/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2614 - mae: 0.3922 - mse: 0.2614 - val_loss: 0.3367 - val_mae: 0.3903 - val_mse: 0.3367\n",
      "Epoch 124/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.3151 - mae: 0.4422 - mse: 0.3151 - val_loss: 0.3327 - val_mae: 0.3886 - val_mse: 0.3327\n",
      "Epoch 125/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.3039 - mae: 0.4445 - mse: 0.3039 - val_loss: 0.3289 - val_mae: 0.3872 - val_mse: 0.3289\n",
      "Epoch 126/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2517 - mae: 0.3896 - mse: 0.2517 - val_loss: 0.3255 - val_mae: 0.3858 - val_mse: 0.3255\n",
      "Epoch 127/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 0.2962 - mae: 0.4203 - mse: 0.2962 - val_loss: 0.3221 - val_mae: 0.3845 - val_mse: 0.3221\n",
      "Epoch 128/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2475 - mae: 0.3878 - mse: 0.2475 - val_loss: 0.3189 - val_mae: 0.3834 - val_mse: 0.3189\n",
      "Epoch 129/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2424 - mae: 0.3880 - mse: 0.2424 - val_loss: 0.3161 - val_mae: 0.3827 - val_mse: 0.3161\n",
      "Epoch 130/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2480 - mae: 0.3821 - mse: 0.2480 - val_loss: 0.3131 - val_mae: 0.3819 - val_mse: 0.3131\n",
      "Epoch 131/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2655 - mae: 0.4121 - mse: 0.2655 - val_loss: 0.3104 - val_mae: 0.3812 - val_mse: 0.3104\n",
      "Epoch 132/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2472 - mae: 0.3912 - mse: 0.2472 - val_loss: 0.3077 - val_mae: 0.3804 - val_mse: 0.3077\n",
      "Epoch 133/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2999 - mae: 0.4093 - mse: 0.2999 - val_loss: 0.3053 - val_mae: 0.3797 - val_mse: 0.3053\n",
      "Epoch 134/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2725 - mae: 0.4163 - mse: 0.2725 - val_loss: 0.3031 - val_mae: 0.3793 - val_mse: 0.3031\n",
      "Epoch 135/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 0.2493 - mae: 0.3872 - mse: 0.2493 - val_loss: 0.3010 - val_mae: 0.3788 - val_mse: 0.3010\n",
      "Epoch 136/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2392 - mae: 0.3805 - mse: 0.2392 - val_loss: 0.2989 - val_mae: 0.3783 - val_mse: 0.2989\n",
      "Epoch 137/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2423 - mae: 0.3881 - mse: 0.2423 - val_loss: 0.2970 - val_mae: 0.3779 - val_mse: 0.2970\n",
      "Epoch 138/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.3137 - mae: 0.4350 - mse: 0.3137 - val_loss: 0.2950 - val_mae: 0.3774 - val_mse: 0.2950\n",
      "Epoch 139/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2741 - mae: 0.4240 - mse: 0.2741 - val_loss: 0.2932 - val_mae: 0.3770 - val_mse: 0.2932\n",
      "Epoch 140/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.3002 - mae: 0.4436 - mse: 0.3002 - val_loss: 0.2919 - val_mae: 0.3766 - val_mse: 0.2919\n",
      "Epoch 141/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2312 - mae: 0.3609 - mse: 0.2312 - val_loss: 0.2904 - val_mae: 0.3762 - val_mse: 0.2904\n",
      "Epoch 142/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2416 - mae: 0.3929 - mse: 0.2416 - val_loss: 0.2891 - val_mae: 0.3758 - val_mse: 0.2891\n",
      "Epoch 143/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2389 - mae: 0.3857 - mse: 0.2389 - val_loss: 0.2879 - val_mae: 0.3755 - val_mse: 0.2879\n",
      "Epoch 144/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2238 - mae: 0.3721 - mse: 0.2238 - val_loss: 0.2866 - val_mae: 0.3751 - val_mse: 0.2866\n",
      "Epoch 145/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2575 - mae: 0.4023 - mse: 0.2575 - val_loss: 0.2853 - val_mae: 0.3746 - val_mse: 0.2853\n",
      "Epoch 146/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2366 - mae: 0.3677 - mse: 0.2366 - val_loss: 0.2843 - val_mae: 0.3746 - val_mse: 0.2843\n",
      "Epoch 147/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2506 - mae: 0.3914 - mse: 0.2506 - val_loss: 0.2833 - val_mae: 0.3747 - val_mse: 0.2833\n",
      "Epoch 148/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2351 - mae: 0.3834 - mse: 0.2351 - val_loss: 0.2825 - val_mae: 0.3748 - val_mse: 0.2825\n",
      "Epoch 149/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2174 - mae: 0.3518 - mse: 0.2174 - val_loss: 0.2817 - val_mae: 0.3748 - val_mse: 0.2817\n",
      "Epoch 150/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2123 - mae: 0.3628 - mse: 0.2123 - val_loss: 0.2810 - val_mae: 0.3749 - val_mse: 0.2810\n",
      "Epoch 151/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2021 - mae: 0.3503 - mse: 0.2021 - val_loss: 0.2802 - val_mae: 0.3750 - val_mse: 0.2802\n",
      "Epoch 152/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2245 - mae: 0.3813 - mse: 0.2245 - val_loss: 0.2796 - val_mae: 0.3750 - val_mse: 0.2796\n",
      "Epoch 153/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2321 - mae: 0.3810 - mse: 0.2321 - val_loss: 0.2790 - val_mae: 0.3751 - val_mse: 0.2790\n",
      "Epoch 154/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2135 - mae: 0.3667 - mse: 0.2135 - val_loss: 0.2784 - val_mae: 0.3752 - val_mse: 0.2784\n",
      "Epoch 155/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2144 - mae: 0.3651 - mse: 0.2144 - val_loss: 0.2779 - val_mae: 0.3752 - val_mse: 0.2779\n",
      "Epoch 156/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2483 - mae: 0.3825 - mse: 0.2483 - val_loss: 0.2775 - val_mae: 0.3753 - val_mse: 0.2775\n",
      "Epoch 157/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2309 - mae: 0.3801 - mse: 0.2309 - val_loss: 0.2770 - val_mae: 0.3754 - val_mse: 0.2770\n",
      "Epoch 158/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.1934 - mae: 0.3409 - mse: 0.1934 - val_loss: 0.2767 - val_mae: 0.3756 - val_mse: 0.2767\n",
      "Epoch 159/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2087 - mae: 0.3634 - mse: 0.2087 - val_loss: 0.2763 - val_mae: 0.3758 - val_mse: 0.2763\n",
      "Epoch 160/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2087 - mae: 0.3637 - mse: 0.2087 - val_loss: 0.2760 - val_mae: 0.3760 - val_mse: 0.2760\n",
      "Epoch 161/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.1966 - mae: 0.3586 - mse: 0.1966 - val_loss: 0.2757 - val_mae: 0.3762 - val_mse: 0.2757\n",
      "Epoch 162/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.1996 - mae: 0.3542 - mse: 0.1996 - val_loss: 0.2755 - val_mae: 0.3764 - val_mse: 0.2755\n",
      "Epoch 163/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.1978 - mae: 0.3598 - mse: 0.1978 - val_loss: 0.2753 - val_mae: 0.3766 - val_mse: 0.2753\n",
      "Epoch 164/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2292 - mae: 0.3768 - mse: 0.2292 - val_loss: 0.2751 - val_mae: 0.3768 - val_mse: 0.2751\n",
      "Epoch 165/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2078 - mae: 0.3592 - mse: 0.2078 - val_loss: 0.2749 - val_mae: 0.3770 - val_mse: 0.2749\n",
      "Epoch 166/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.1953 - mae: 0.3471 - mse: 0.1953 - val_loss: 0.2748 - val_mae: 0.3771 - val_mse: 0.2748\n",
      "Epoch 167/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2318 - mae: 0.3737 - mse: 0.2318 - val_loss: 0.2747 - val_mae: 0.3773 - val_mse: 0.2747\n",
      "Epoch 168/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2414 - mae: 0.3867 - mse: 0.2414 - val_loss: 0.2746 - val_mae: 0.3775 - val_mse: 0.2746\n",
      "Epoch 169/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.1921 - mae: 0.3409 - mse: 0.1921 - val_loss: 0.2745 - val_mae: 0.3777 - val_mse: 0.2745\n",
      "Epoch 170/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2122 - mae: 0.3639 - mse: 0.2122 - val_loss: 0.2745 - val_mae: 0.3779 - val_mse: 0.2745\n",
      "Epoch 171/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2263 - mae: 0.3811 - mse: 0.2263 - val_loss: 0.2744 - val_mae: 0.3782 - val_mse: 0.2744\n",
      "Epoch 172/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2460 - mae: 0.3887 - mse: 0.2460 - val_loss: 0.2744 - val_mae: 0.3784 - val_mse: 0.2744\n",
      "Epoch 173/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2213 - mae: 0.3681 - mse: 0.2213 - val_loss: 0.2744 - val_mae: 0.3786 - val_mse: 0.2744\n",
      "Epoch 174/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2245 - mae: 0.3648 - mse: 0.2245 - val_loss: 0.2744 - val_mae: 0.3788 - val_mse: 0.2744\n",
      "Epoch 175/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2188 - mae: 0.3628 - mse: 0.2188 - val_loss: 0.2744 - val_mae: 0.3791 - val_mse: 0.2744\n",
      "Epoch 176/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2068 - mae: 0.3638 - mse: 0.2068 - val_loss: 0.2744 - val_mae: 0.3793 - val_mse: 0.2744\n",
      "Epoch 177/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2108 - mae: 0.3513 - mse: 0.2108 - val_loss: 0.2745 - val_mae: 0.3795 - val_mse: 0.2745\n",
      "Epoch 178/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.1750 - mae: 0.3227 - mse: 0.1750 - val_loss: 0.2745 - val_mae: 0.3796 - val_mse: 0.2745\n",
      "Epoch 179/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.1966 - mae: 0.3493 - mse: 0.1966 - val_loss: 0.2745 - val_mae: 0.3798 - val_mse: 0.2745\n",
      "Epoch 180/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2162 - mae: 0.3623 - mse: 0.2162 - val_loss: 0.2746 - val_mae: 0.3800 - val_mse: 0.2746\n",
      "Epoch 181/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2102 - mae: 0.3494 - mse: 0.2102 - val_loss: 0.2747 - val_mae: 0.3803 - val_mse: 0.2747\n",
      "Epoch 182/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2252 - mae: 0.3683 - mse: 0.2252 - val_loss: 0.2747 - val_mae: 0.3804 - val_mse: 0.2747\n",
      "Epoch 183/250\n",
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 0.2160 - mae: 0.3690 - mse: 0.2160 - val_loss: 0.2748 - val_mae: 0.3806 - val_mse: 0.2748\n",
      "Epoch 183: early stopping\n",
      "\n",
      "Duration :  00:00:10 260ms\n"
     ]
    }
   ],
   "source": [
    "vID.chrono_start()\n",
    "ANNhistory = ANNmodel.fit(x_train,\n",
    "                    y_train,\n",
    "                    epochs          = 250,\n",
    "                    batch_size      = 10,\n",
    "                    verbose         = 1,\n",
    "                    validation_data = (x_test, y_test),\n",
    "                    callbacks=[es])\n",
    "vID.chrono_show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ef58b2ec-896c-4dfc-b20f-7b4cf3bc4cb3",
   "metadata": {},
   "source": [
    "#### Accuracy of the model\n",
    "\n",
    "<div class=\"intro\">\n",
    "    \n",
    "The accurracy will now be evaluated by calculating the MAE, MSE and $R^2$ quantities\n",
    "\n",
    "MAE =  Mean Absolute Error \n",
    "$$\\mathrm{MAE}=\\frac{1}{n}\\sum_{i}\\left|\\hat{Y}(x_{i})-Y(x_{i})\\right|$$\n",
    "\n",
    "MSE = Mean Squared Error  \n",
    "$$\\mathrm{MSE}=\\frac{1}{n}\\sum_{i}\\left(\\hat{Y}(x_{i})-Y(x_{i})\\right)^{2}$$\n",
    "\n",
    "$R^2$ = Coefficient of Determination\n",
    "$$R^{2}=1-\\frac{\\sum_{i}\\left(Y(x_{i})-\\hat{Y}(x_{i})\\right)^{2}}{\\sum_{i}\\left(Y(x_{i})-\\bar{Y}\\right)^{2}}$$\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1c1947f4-7ee1-46e2-8d60-daa88d3755cd",
   "metadata": {},
   "source": [
    "##### Accurracy of the optimized (trained) model\n",
    "\n",
    "<div class=\"intro\">\n",
    "\n",
    "MAE and MSE are returned by the model, given the `compile` command:\n",
    "\n",
    "```python\n",
    "model.compile(optimizer = 'adam',\n",
    "              loss      = 'mse',\n",
    "              metrics   = ['mae', 'mse'] )\n",
    "```\n",
    "<br>\n",
    "\n",
    "They can be obtained with the `evaluate` command.\n",
    "\n",
    "$R^2$ must be calculated separately. It involves to apply the percepron model on the training and test sets. This is done with the `predict` command. <br>\n",
    "`predict` returns an array of $\\hat{y}$ predicted values.\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "49f58175-bcfc-415f-a757-205560a05227",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[92mx_train / loss      : 0.2154\n",
      "x_train / mae       : 0.3639\n",
      "x_train / mse       : 0.2154\n",
      "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step \n",
      "x_train / R^2       : 0.9304\u001b[0m\n",
      "\n",
      "\u001b[94mx_test / loss       : 0.2748\n",
      "x_test / mae        : 0.3806\n",
      "x_test / mse        : 0.2748\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 18ms/step\n",
      "x_test / R^2       : 0.9085\u001b[0m\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import r2_score\n",
    "\n",
    "EvTrain = ANNmodel.evaluate(x_train, y_train, verbose=0)\n",
    "print(f\"{fg.GREEN}x_train / loss      : {EvTrain[0]:5.4f}\")\n",
    "print(f'x_train / mae       : {EvTrain[1]:5.4f}')\n",
    "print(f\"x_train / mse       : {EvTrain[2]:5.4f}\")\n",
    "ytrain_hat = ANNmodel.predict(x_train)\n",
    "R2_train = r2_score(y_train,ytrain_hat)\n",
    "print(f\"x_train / R^2       : {R2_train:5.4f}{fg.OFF}\")\n",
    "print()\n",
    "\n",
    "\n",
    "EvTest = ANNmodel.evaluate(x_test, y_test, verbose=0)\n",
    "print(f\"{fg.BLUE}x_test / loss       : {EvTest[0]:5.4f}\")\n",
    "print(f'x_test / mae        : {EvTest[1]:5.4f}')\n",
    "print(f\"x_test / mse        : {EvTest[2]:5.4f}\")\n",
    "ytest_hat = ANNmodel.predict(x_test)\n",
    "R2_test = r2_score(y_test,ytest_hat)\n",
    "print(f\"x_test / R^2       : {R2_test:5.4f}{fg.OFF}\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f8271f33-fa55-4788-a412-259e68978b62",
   "metadata": {},
   "source": [
    "##### Accurracy as a function of epochs\n",
    "\n",
    "<div class=\"intro\">\n",
    "\n",
    "Plotting the loss function or the MAE as a function of epochs is a common practice in machine learning. It serves several purposes and provides insights into the training process:\n",
    "\n",
    "- **model convergence**: the goal is for the loss to decrease over time. A decreasing loss indicates that the model is learning and improving its ability to make accurate predictions on the training data\n",
    "- **overfitting and underfitting**:\n",
    "    - if the loss on the training data continues to decrease while the loss on a validation set (data the model has not seen during training) starts increasing, it might indicate **overfitting**. It occurs when the model learns the training data too well, **including its noise and outliers**, and fails to generalize to new, unseen data.\n",
    "    - If the loss is still high after many epochs, it might indicate **underfitting**, *i.e.* the model is too simple to capture the underlying patterns in the data\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "28d9e2ad-0b79-46ff-b95b-d68c6f572d54",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f21044afb90>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1400x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1400x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rcParams[\"figure.figsize\"] = (14,6)\n",
    "\n",
    "df=pd.DataFrame(data=ANNhistory.history)\n",
    "figLOSS=df.plot(y=[\"loss\",\"val_loss\"],linestyle='-', marker='o',fontsize=14)\n",
    "figLOSS.set_xlabel('epoch',fontdict={'fontsize':16})\n",
    "figLOSS.set_ylabel('loss',fontdict={'fontsize':16})\n",
    "figLOSS.set_ylim([0.0,10.0])\n",
    "figLOSS.legend(loc='upper right', shadow=True, fontsize='x-large')\n",
    "\n",
    "figMAE=df.plot(y=[\"mae\",\"val_mae\"],linestyle='-', marker='o',fontsize=14)\n",
    "figMAE.set_xlabel('epoch',fontdict={'fontsize':16})\n",
    "figMAE.set_ylabel('mae',fontdict={'fontsize':16})\n",
    "figMAE.set_ylim([0.0,4.0])\n",
    "figMAE.legend(loc='upper right', shadow=True, fontsize='x-large')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "62289823-8c22-42a8-a26b-3f545ae8ace6",
   "metadata": {},
   "source": [
    "<div class='rq'>\n",
    "\n",
    "There is a significant improvement after several tenths of epochs. It's quite a long process for such a simple regression. \n",
    "\n",
    "Most of the time, the early stopping is not activated, the training still slightly improves. \n",
    "\n",
    "Unless you were unlucky, the training converged well and there is a nice similar performance of the model on both the training set and the validation set. This is the sign that there is no apparent underfitting or overfitting.\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "37cec342-5c6d-440d-9999-344bd4372fd3",
   "metadata": {},
   "source": [
    "##### Prediction of the petal length of any iris as a function of its width\n",
    "\n",
    "<div class=\"intro\">\n",
    "\n",
    "Let's now plot the $\\hat{y}_\\mathrm{train}$ and $\\hat{y}_\\mathrm{test}$ values as a function of their actual counterpart. This is a standard way to visually assess the performance of the model. The residual will also be plotted. It is a well-known property used in statistics. It is the difference between the observed value and the estimated value of the quantity of interest. They provide a diagnostic tool to identify potential issues: if there is a **systematic pattern** in the residuals (e.g., a curve or specific shape), it suggests that **the model may be missing some important relationship**. Unusual or extreme values in the residuals may indicate **outliers** in the data.\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "67b0fa31-2bf0-4a23-975c-5ba0392659f5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div style='text-align:center; font-weight: bold; font-size:16px;background-color: #343132;color: #ffffff'>Basic plots</div>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x1000 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x1000 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "t4pPC.centerTitle(\"Basic plots\")\n",
    "plt.rcParams[\"figure.figsize\"] = (10,10)\n",
    "\n",
    "maxL= 8.0\n",
    "figP, axP = plt.subplots()\n",
    "axP.plot(y_train,ytrain_hat,'ro',label='training set')\n",
    "axP.plot(y_test,ytest_hat,'b+',label='validation set')\n",
    "axP.legend(loc='upper left', shadow=True, fontsize='x-large')\n",
    "axP.set_xlim(0.0,maxL); axP.set_ylim(0.0,maxL)\n",
    "axP.set_xlabel('actual petal length',fontsize=16)\n",
    "axP.set_ylabel('predicted petal length',fontsize=16)\n",
    "axP.tick_params(labelsize = 14)\n",
    "\n",
    "diffyt = ytrain_hat - y_train\n",
    "diffyp = ytest_hat - y_test\n",
    "figE, axE = plt.subplots()\n",
    "axE.plot(y_train,diffyt,'ro',label='training set')\n",
    "axE.plot(y_test,diffyp,'b+',label='validation set')\n",
    "axE.legend(loc='upper left', shadow=True, fontsize='x-large')\n",
    "axE.set_xlim(0.0,maxL); axE.set_ylim(-2,2)\n",
    "axE.set_xlabel('actual petal length',fontsize=16)\n",
    "axE.set_ylabel('residual',fontsize=16)\n",
    "axE.tick_params(labelsize = 14)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "20319f82-ad5e-4f41-b611-a42f35825149",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div style='text-align:center; font-weight: bold; font-size:16px;background-color: #343132;color: #ffffff'>Fancy plots with seaborn</div>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[92m Train. mean: -0.038    std: 0.463    MAE: 0.364     R2: 0.965\n",
      "\u001b[94m  Test. mean: -0.021    std: 0.524    MAE: 0.381     R2: 0.956\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Error distribution</th>\n",
       "      <th>Type</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.193852</td>\n",
       "      <td>Test set</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.240531</td>\n",
       "      <td>Test set</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.017191</td>\n",
       "      <td>Test set</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.206148</td>\n",
       "      <td>Test set</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.106148</td>\n",
       "      <td>Test set</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>145</th>\n",
       "      <td>-0.862699</td>\n",
       "      <td>Training set</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>146</th>\n",
       "      <td>-0.426075</td>\n",
       "      <td>Training set</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>147</th>\n",
       "      <td>-1.186038</td>\n",
       "      <td>Training set</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>148</th>\n",
       "      <td>0.093852</td>\n",
       "      <td>Training set</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149</th>\n",
       "      <td>0.083980</td>\n",
       "      <td>Training set</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>150 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Error distribution          Type\n",
       "0              0.193852      Test set\n",
       "1              0.240531      Test set\n",
       "2              0.017191      Test set\n",
       "3             -0.206148      Test set\n",
       "4             -0.106148      Test set\n",
       "..                  ...           ...\n",
       "145           -0.862699  Training set\n",
       "146           -0.426075  Training set\n",
       "147           -1.186038  Training set\n",
       "148            0.093852  Training set\n",
       "149            0.083980  Training set\n",
       "\n",
       "[150 rows x 2 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_7366/1614879346.py:46: FutureWarning: \n",
      "\n",
      "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `y` variable to `hue` and set `legend=False` for the same effect.\n",
      "\n",
      "  fig2=sns.boxplot(data=distrib, x=\"Error distribution\", y=\"Type\", notch=True, orient = 'h',palette='Spectral')\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[Text(0, 0.5, '')]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x1300 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "t4pPC.centerTitle(\"Fancy plots with seaborn\")\n",
    "\n",
    "import seaborn as sns\n",
    "from matplotlib import gridspec\n",
    "# print(diffyt.values)\n",
    "# print(diffyt.values.ravel())\n",
    "print(f\"{fg.GREEN}\",\n",
    "      f\"Train. mean: {np.mean(diffyt.values.ravel()):6.3f}\",\n",
    "      f\"   std: {np.std(diffyt.values.ravel()):.3f}\",\n",
    "      f\"   MAE: {np.average(abs(diffyt.values.ravel())):.3f}\",\n",
    "      f\"    R2: {np.corrcoef(y_train.values.ravel(),ytrain_hat.ravel())[0,1]:.3f}\"\n",
    "     )\n",
    "print(f\"{fg.BLUE}\",\n",
    "      f\" Test. mean: {np.mean(diffyp.values.ravel()):6.3f}\",\n",
    "      f\"   std: {np.std(diffyp.values.ravel()):.3f}\",\n",
    "      f\"   MAE: {np.average(abs(diffyp.values.ravel())):.3f}\",\n",
    "      f\"    R2: {np.corrcoef(y_test.values.ravel(),ytest_hat.ravel())[0,1]:.3f}\"\n",
    "     )\n",
    "\n",
    "# make a dataframe\n",
    "df1 = pd.DataFrame(\n",
    "    {\n",
    "        \"Error distribution\": diffyp.values.ravel(),\n",
    "        \"Type\": 'Test set'    },\n",
    ")\n",
    "df2 = pd.DataFrame(\n",
    "    {\n",
    "        \"Error distribution\": diffyt.values.ravel(),\n",
    "        \"Type\": 'Training set'    },\n",
    ")\n",
    "distrib=pd.concat([df1,df2],ignore_index=True)\n",
    "display(distrib)\n",
    "\n",
    "xlim=2.0\n",
    "nbins=20\n",
    "plt.rcParams[\"figure.figsize\"] = (8,13) #on ajuste la taille du graphique\n",
    "plt.rcParams[\"font.size\"] = (18) #on ajuste la taille des caractères\n",
    "gs = gridspec.GridSpec(2, 1, height_ratios=[3, 1])\n",
    "plt.subplot(gs[0])\n",
    "fig1=sns.histplot(data=distrib,x=\"Error distribution\", hue=\"Type\", bins=nbins, kde=True, palette='Spectral', alpha=1.0)\n",
    "plt.xlabel(\"$\\mathrm{petal\\:length}^{\\mathrm{pred}}-\\mathrm{petal\\:length}^{\\mathrm{exp}}$\")\n",
    "plt.ylabel(\"number of data points in each interval\")\n",
    "plt.xlim(-xlim,xlim)\n",
    "plt.subplot(gs[1])\n",
    "plt.xlim(-xlim,xlim)\n",
    "fig2=sns.boxplot(data=distrib, x=\"Error distribution\", y=\"Type\", notch=True, orient = 'h',palette='Spectral')\n",
    "fig2.set(ylabel=None)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a9964eef-85ba-432b-868b-2ba9acdd32d8",
   "metadata": {},
   "source": [
    "### Comparison of the least-square fit linear and perceptron-based linear regressions"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3bb59807-536d-4f12-8416-7a4cbb7a1687",
   "metadata": {},
   "source": [
    "#### Parameters w and b of the trained perceptron model\n",
    "\n",
    "<div class=\"rq\">\n",
    "\n",
    "Given the formula that calculates the output as a function of the input, this is one of the rare cases, if not the only one, where we can directly give meaning to the model parameters.\n",
    "\n",
    "<p style=\"text-align: center\"><img width=\"350px\" src=\"DS4B-svg/1neuron-perceptron-woBackP.svg\" style=\"margin-left:auto; margin-right:auto\" id=\"img_perceptron\"/></p>\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "dfbdaf38-f614-4fa5-acf1-27b8e72f6fa2",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m\u001b[91mOutput Layer\u001b[0m\n",
      "w: [[2.2333944]]\n",
      "b: [1.0471728]\n"
     ]
    }
   ],
   "source": [
    "print(f\"{color.BOLD}{color.RED}Output Layer{color.OFF}\")\n",
    "print(f\"w: {ANNmodel.layers[0].get_weights()[0]}\")\n",
    "print(f\"b: {ANNmodel.layers[0].get_weights()[1]}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5f2be930-d668-4254-ac89-d873934ac867",
   "metadata": {},
   "source": [
    "#### Linear regression with SciPy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "9c77f4f9-f199-4b76-b05d-5068c7092edf",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "a   =    2.226\n",
      "b   =    1.091\n",
      "R^2 =  0.927\n",
      "Error on the slope =   0.051\n",
      "Error on the y-intercept =   0.073\n"
     ]
    }
   ],
   "source": [
    "from scipy.stats import linregress\n",
    "pw=dfi[\"petal_width\"].to_numpy()\n",
    "pl=dfi[\"petal_length\"].to_numpy()\n",
    "fit = linregress(x=pw, y=pl)\n",
    "\n",
    "a = fit.slope\n",
    "b = fit.intercept\n",
    "r = fit.rvalue\n",
    "stderr = fit.stderr\n",
    "istderr = fit.intercept_stderr\n",
    "\n",
    "print(f\"a   = {a:8.3f}\\nb   = {b:8.3f}\\nR^2 = {r**2:6.3f}\")\n",
    "print(f\"Error on the slope ={stderr:8.3f}\\nError on the y-intercept ={istderr:8.3f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e0527709-c7c5-4ff7-ae64-6e0823ca0283",
   "metadata": {},
   "source": [
    "#### Graphical comparison"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "6193e90c-db97-4d76-8a4e-60ccd4ab42f3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f20fcd18310>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x1200 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "def y_hat(a,b,x):\n",
    "    return a * x + b\n",
    "# on calcule pl_hat la longueur prédite, pour chacune des largeurs (pw) mesurées expérimentalement\n",
    "pl_hat = y_hat(a,b,pw)\n",
    "\n",
    "plt.rcParams[\"figure.figsize\"] = (12,12) #on ajuste la taille du graphique\n",
    "plt.plot(pw, pl,linestyle='', marker='o', color='blue', label='observations exp. ($y$)')\n",
    "plt.plot(pw,pl_hat,color='red',label='prediction by lin. regr. ($\\hat{y}_\\mathrm{lin.reg.}$)')\n",
    "plt.plot(x_train, ytrain_hat,linestyle='', marker='P', markersize=20, color='green', label='perceptron prediction ($\\hat{y}_\\mathrm{ANN}(\\mathrm{training})$)')\n",
    "plt.plot(x_test, ytest_hat,linestyle='', marker='X', markersize=20, color='orange', label='perceptron prediction ($\\hat{y}_\\mathrm{ANN}(\\mathrm{test})$)')\n",
    "plt.xlabel(\"petal width\") # label de l'axe des abscisses\n",
    "plt.ylabel(\"petal length\") # label de l'axe des ordonnées\n",
    "plt.legend() #affichage de la légende sur le graphe"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a384e0e0-a499-47ca-bf96-fe3e193799ac",
   "metadata": {},
   "source": [
    "<div class=\"rq\">\n",
    "\n",
    "The linear regression approach is more accurate. But at this scale, there is in general no significant difference between the two results. Yet, given the random component of the training and of the holdout method, the perceptron may fail.\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "10bea743-28dd-460a-a31b-1517267ecf25",
   "metadata": {},
   "source": [
    "## Comments, conclusion and outlook\n",
    "\n",
    "<div class=\"rq\">\n",
    "\n",
    "This case is a nice and simple playground to understand the **basic concepts of supervised learning**.\n",
    "\n",
    "When it comes to 1D linear regression specifically, using a least squares fitting algorithm can often be more computationally efficient and straightforward compared to training a neural network, even a single perceptron.\n",
    "\n",
    "The least squares fitting method directly calculates the optimal values for the slope and intercept of the linear relationship by minimizing the sum of squared differences between the observed and predicted values. This approach does not require an iterative optimization process, making it computationally efficient for linear regression tasks.\n",
    "\n",
    "In contrast, training a neural network, even a simple one like a single perceptron, involves an iterative optimization process using methods like gradient descent. **For simple linear regression problems, the additional complexity of a neural network may not be justified when a solution like least squares is available**.\n",
    "\n",
    "**Yet, for more complex cases, the conclusion would probably not be the same**\n",
    "</div>\n",
    "\n",
    "### The perceptron model with an activation function\n",
    "\n",
    "<div class=\"intro\">\n",
    "\n",
    "This would be a **logistic regression**, that provides a **classification**. The model would provide a **probability** that a given input data point belongs to a **particular class**. It would require to add an **activation function to the output neuron**. The **loss function** would also differ. The function of choice is the so-called **cross entropy**\n",
    "\n",
    "<img src=\"./DS4B-Slides/perceptron8.jpg\" alt=\"perceptron8\" width=\"1200\" style=\"display: block; margin: 0 auto; align:center\"/>\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5fdff0f3-2b7a-40d5-a81e-d952f7b96b4b",
   "metadata": {},
   "source": [
    "### Beyond the perceptron model: deep learning\n",
    "\n",
    "<div class=\"intro\">\n",
    "\n",
    "The scheme below illustrates what is deep learning with **dense layers**, also known as **fully connected layers** or **feedforward layers**. It  is a fundamental aspect of neural network architectures. Dense layers are commonly used to model complex relationships in data, making them a crucial component of deep neural network. \n",
    "\n",
    "</div>\n",
    "<br>\n",
    "<div class=\"introT\" title=\"Dense layer basics\">\n",
    "\n",
    "- **Neurons and Activation Functions:**\n",
    "    - A dense layer consists of **neurons** or **nodes**, each associated with a **weight** and a **bias**\n",
    "    - Neurons apply a **weighted sum** of the input data along with a bias term\n",
    "    - The result of this sum is often passed through an **activation function**, introducing non-linearity into the model.\n",
    "\n",
    "\n",
    "- **Weights and Biases:**\n",
    "    - Each neuron in a dense layer is connected to **every neuron** in the previous layer. These connections are represented by weights.\n",
    "    - The **weights and biases** are parameters **learned during the training process** to optimize the model for a specific task.\n",
    "\n",
    "\n",
    "- **Input and Output:** \n",
    "    - The **input to a dense layer** is a **vector** (or tensor) of features from the **previous layer**.\n",
    "    - The **output** is a **transformed representation** of the **input**.\n",
    "\n",
    "<img src=\"./DS4B-Slides/perceptron9.jpg\" alt=\"perceptron9\" width=\"1200\" style=\"display: block; margin: 0 auto; align:center\"/>\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "37aa1455-a21c-46f5-a67c-814691f11116",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true
   },
   "source": [
    "### Architectures of Artificial Neural Networks (ANN): *the neural network zoo*\n",
    "\n",
    "<div class=\"intro\">\n",
    "\n",
    "With new neural network architectures popping up every now and then, it’s hard to keep track of them all. Knowing all the abbreviations can also be a bit overwhelming at first. There is very nice cheat sheet on the web, that contains many of those architectures. Most of these are neural networks, some are completely different beasts. This cheat sheet also shows the effort that has to be made to learn these techniques.\n",
    "\n",
    "[Van Veen, F. & Leijnen, S. (2019). The Neural Network Zoo. © Asimov Institute, 2019](https://www.asimovinstitute.org/neural-network-zoo/)\n",
    "\n",
    "<img src=\"./DS4B-Slides/pngs/NeuralNetworkZoo20042019.png\" alt=\"NNZoo\" width=\"800\" style=\"display: block; margin: 0 auto; align:center\"/>\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9c63f492-816b-41f3-b4f3-5cd753c35813",
   "metadata": {},
   "source": [
    "### Is it possible to do machine learning with something else than ANNs?\n",
    "\n",
    "<div class=\"intro\">\n",
    "\n",
    "[Yes! see for example \"Top 10 Machine Learning Algorithms For Beginners: Supervised, and More\"](https://www.simplilearn.com/10-algorithms-machine-learning-engineers-need-to-know-article)\n",
    "\n",
    "<img src=\"./DS4B-Slides/perceptron10.jpg\" alt=\"perceptron11\" width=\"1200\" style=\"display: block; margin: 0 auto; align:center\"/>\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "36f3d566-303a-46b5-9341-ea82ae44bfa6",
   "metadata": {},
   "source": [
    "### Prerequisite to do ML on your own\n",
    "\n",
    "<div class=\"intro\">\n",
    "\n",
    "<img src=\"./DS4B-Slides/perceptron11.jpg\" alt=\"perceptron11\" width=\"1200\" style=\"display: block; margin: 0 auto; align:center\"/>\n",
    "<br>\n",
    "<br>\n",
    "<img src=\"./DS4B-Slides/perceptron12.jpg\" alt=\"perceptron12\" width=\"1200\" style=\"display: block; margin: 0 auto; align:center\"/>\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cab9435a-1af2-465e-b7a8-dcd633f84acd",
   "metadata": {},
   "source": [
    "# Useful links\n",
    "\n",
    "[Wikipedia page on the perceptron model](https://en.wikipedia.org/wiki/Perceptron)\n",
    "\n",
    "[Hands-on machine learning with scikit-learn, keras and tensorflow (3rd edition, **2022**), Aurélien Géron, O'Reilly editions](https://github.com/ageron/handson-ml3)\n",
    "\n",
    "[TensorFlow 2 Tutorial: Get Started in Deep Learning with tf.keras](https://machinelearningmastery.com/tensorflow-tutorial-deep-learning-with-tf-keras/)\n",
    "\n",
    "Some images were generated with [DALL·E](https://openai.com/dall-e-2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "4dea101f-ce6f-47ee-819c-a2fb77f6f708",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/markdown": [
       "**End at:** Wednesday 27 March 2024, 12:26:09  \n",
       "**Duration:** 00:06:30 896ms"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/markdown": [
       "<p style=\"text-align: center\"><img width=\"800px\" src=\"./config/svg/logoEnd.svg\" style=\"margin-left:auto; margin-right:auto\"/></p>"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "vID.end(cwd0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b6af59d5-ee74-417c-87d0-2cc5096025c6",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.9"
  },
  "toc-autonumbering": false,
  "toc-showmarkdowntxt": false
 },
 "nbformat": 4,
 "nbformat_minor": 5
}