regression-tf-notebook.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Untitled0.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "Vkkgm55UgDBb",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import tensorflow as tf\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "from pylab import rcParams\n",
        "import matplotlib.pyplot as plt\n",
        "import warnings"
      ],
      "execution_count": 118,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "w1C8jlMNkowB",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "%matplotlib inline\n",
        "warnings.filterwarnings(\"ignore\")\n",
        "pd.set_option('display.expand_frame_repr', False)\n",
        "rcParams['figure.figsize'] = 16, 8"
      ],
      "execution_count": 119,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "McV3mBetgtU8",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df = pd.read_csv(\"data.csv\")"
      ],
      "execution_count": 120,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "T0te_PCjg6L7",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 202
        },
        "outputId": "fd16c418-4d08-49f9-a615-f35b36709b0c"
      },
      "source": [
        "df.head()"
      ],
      "execution_count": 121,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>X</th>\n",
              "      <th>y</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>230.1</td>\n",
              "      <td>22.1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>44.5</td>\n",
              "      <td>10.4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>17.2</td>\n",
              "      <td>9.3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>151.5</td>\n",
              "      <td>18.5</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>180.8</td>\n",
              "      <td>12.9</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       X     y\n",
              "0  230.1  22.1\n",
              "1   44.5  10.4\n",
              "2   17.2   9.3\n",
              "3  151.5  18.5\n",
              "4  180.8  12.9"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 121
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nw12IzEOnTG_",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 483
        },
        "outputId": "84545186-917e-4c87-b18f-1b7dc4b7258c"
      },
      "source": [
        "plt.scatter(df[\"X\"], df[\"y\"])\n",
        "plt.show()"
      ],
      "execution_count": 122,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA54AAAHSCAYAAACEihTZAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3db4xl6Vkg9ueddrG5XhBl4onlqcWZ0cYqy2yLqaWEiHoV2U2WZvEmLhpC4AOLEiuzH0ACZJVSw4fgVSJ1JR2wFCVBMcLCWe2CnfVQOAzajuWeFdpWwm41ZWj/a+HsGIU7gz2bnQI2U2Lb7Tcf6lZ3V/X9X+f/+f2kUXefujXnrXPfc+o+53nO86accwAAAEBZnqh7AAAAAHSbwBMAAIBSCTwBAAAolcATAACAUgk8AQAAKJXAEwAAgFK9qcqdvfWtb81PP/10lbsEAACgIrdv3/6XOecnz26vNPB8+umnY39/v8pdAgAAUJGU0h+N267UFgAAgFIJPAEAACiVwBMAAIBSCTwBAAAolcATAACAUgk8AQAAKJXAEwAAgFIJPAEAACiVwBMAAIBSCTwBAAAolcATAACAUgk8AQAAKJXAEwAAgFIJPAEAACiVwBMAAIBSvanuAQAAABRl72AY12/cjVcOj+Kp1UFsX1mPrY21uofVewJPAACgE/YOhvH8C3fi6N79iIgYHh7F8y/ciYgQfNZMqS0AANAJ12/cfRB0nji6dz+u37hb04g4IfAEAAA64ZXDo4W2Ux2BJwAA0AlPrQ4W2k51BJ4AAEAnbF9Zj8HKhVPbBisXYvvKek0j4oTmQgAAQCecNBDS1bZ5BJ4AAEBnbG2sCTQbSKktAAAApRJ4AgAAUCqBJwAAAKUSeAIAAFAqzYUAAADOYe9gqJPuDAJPAACAJe0dDOP5F+7E0b37ERExPDyK51+4ExEh+HyEUlsAAIAlXb9x90HQeeLo3v24fuNuTSNqJoEnAADAkl45PFpoe18JPAEAAJb01Opgoe19JfAEAABY0vaV9RisXDi1bbByIbavrNc0ombSXAgAAGBJJw2EdLWdTuAJAABwDlsbawLNGZTaAgAAUCqBJwAAAKUSeAIAAFAqz3gCAACcsXcw1DCoQAJPAACAR+wdDOP5F+7E0b37ERExPDyK51+4ExEh+FySwBMAAEoia9bOY3D9xt0HQeeJo3v34/qNuwuNvY0/e1kEngAAUAJZs/Yeg1cOjxbaPk5bf/ayaC4EAAAlmJY164u2HoOnVgcLbR+nrT97WQSeAABQgiKyZm3X1mOwfWU9BisXTm0brFyI7Svrc/8/2vqzl0XgCQAAJSgia9Z2bT0GWxtrce3qxVhbHUSKiLXVQVy7enGhEtm2/uxlmRl4ppS+PaX0UkrpCymlz6eUfnq0/UMppWFK6bOj/36g/OECAEA7FJE1a7s2H4OtjbW4tXM5Xt59X9zaubzwc5lt/tnLME9zoa9HxAdzzr+XUvqWiLidUvr06Gsfzjn/9+UNDwAA2ukkUOlzV9M+H4M+/+zjpJzzYt+Q0m9GxP8YEZci4l8vEnhubm7m/f39xUYIAABAK6SUbuecN89uX+gZz5TS0xGxERG/O9r0UymlP0gpfTSl9JZzjxIAAIDOmTvwTCl9c0R8MiJ+Juf8ZxHxSxHxVyPi2Yh4NSJ+YcL3PZdS2k8p7b/22msFDBkAAIA2mecZz0gprcRx0PkPcs4vRETknL/6yNd/OSJ+a9z35pw/EhEfiTgutT3vgAEAaJe9g6Hn3KDnZgaeKaUUEb8SEV/MOf/iI9vfnnN+dfTPH4yIz5UzRAAA2mrvYBjPv3Anju7dj4iI4eFRPP/CnYgIwSf0yDyltpci4scj4vKZpVP+u5TSnZTSH0TEeyPiZ8scKAAA7XP9xt0HQeeJo3v34/qNuzWNCKjDzIxnzvmfRkQa86XfLn44AAB0ySuHRwttB7ppoa62AACwiKdWBwttB7pJ4AkAQGm2r6zHYOXCqW2DlQuxfWW9phHBaXsHw7i0ezOe2XkxLu3ejL2DYd1D6qS5utoCAMAyThoI6WpLE2l+VR2BJwAApdraWPMhnkaa1vzKnC2WwBMAAJiqq2uxan5VHc94AgAAE52Uow4PjyLHw3LULjwLqflVdQSeAADARF1ei1Xzq+ootQUAACbqcjmq5lfVEXgCAAATPbU6iOGYILMr5aiaX1VDqS0AADCRclSKIOMJANBTXe1USrGUo9avC+eqwBMAoIdOOpWeNI056VQaEa37QEv5lKPWpyvnqlJbAICa7B0M49LuzXhm58W4tHuz0uUputypFLqkK+eqjCcAQA3qzmJ0uVMpdElXzlUZTwCAGtSdxZjUkbQrnUqhK7pyrgo8AQBqUHcWQ6dSaIeunKtKbQEAalD32og6lUI7dOVcTTnnyna2ubmZ9/f3K9sfAEBTnX3GM+I4i3Ht6sXWfaAEOJFSup1z3jy7XcYTAKAGXcliAMxD4AkAUBNrIwJ9IfAEAIAl7B0MZaxhTgJPAABYUN3rsELbWE4FAAAWVPc6rNA2Ak8AAFhQ3euwQtsIPAEAYEGT1lutah1WaBuBJwAALGj7ynoMVi6c2jZYuRDbV9ZrGhE0m+ZCAACwIOuwwmIEngAAsATrsML8lNoCAABQKhlPAICW2jsY1lrqWff+gfYQeAIAtNDewTCef+HOg7Ukh4dH8fwLdyIiKgn+6t4/0C5KbQEAWuj6jbsPgr4TR/fux/Ubd3uxf6BdZDwBAApWRQnqK4dHC20vWt37B9pFxhMAoEAnJajDw6PI8bAEde9gWOh+nlodLLS9aHXvH2gXgScAQIGqKkHdvrIeg5ULp7YNVi7E9pX1QvezyP5TRLz3XU9Wsn+gXQSeAAAFqqoEdWtjLa5dvRhrq4NIEbG2OohrVy9W1thna2Mtfui71iI9si1HxCdvDwvP7nLa3sEwLu3ejGd2XoxLuzcdb1rBM54AAAV6anUQwzFBZhklqFsba7V2kH3pS69FPrPtJLurs205dBOmrWQ8AQAKVHcJbJWa2GCo69lA3YRpKxlPAIACnWSdyu5q2wRVZnfn0YdsYBODfZiHwBMAoGB1l8BWZfvK+qlAL6Le7O60bGBX3o+mBfswL6W2AAAspe4GR2f1IRvYp1JuukXGEwCApTUpu9uHbGCfSrnpFoEnAACd0ITS372DYelBYZOCfZiXwBMAgLGqCKKKVHc2sA/NjWBZAk8AAB7T1iCqzmxgH5obwbI0FwIA4DHWi1xcH5obwbJkPAEAeEzTg6gmlgH3obkRLEvGEwCAx0wKlpoQRJ2UAQ8PjyLHwzLgvYNhreOy1AlMJvAEAOAxTQ6imloG3LR1TWmHvYNhXNq9Gc/svBiXdm/WfgOlLEptAQB4TN0dYqdpchmwpU5YRFubeC1D4AkAwFhNDaI8S0lX9KkTslJbAKAT+lKuRrPLgGERTc7eF03GEwBovT6Vq9HsMmBYRJ+y9wJPAKD1+lSuxrGmlgHDIravrJ+6aRbR3ey9wBMAaL0+lasB3dGn7L3AEwBovT6VqwHd0pfsveZCAEDraTYD0GwyngBA6/WpXA36bO9g6DxvKYEnANAJfSlXg77SvbrdBJ4AQKfJkFAE86h+Xe5e3Yf5JfAEADpLhoQimEfN0NXu1X2ZX5oLAQCdNS1DAvMyj5phUpfqtnev7sv8EngCAJ3V1QwJ1TKPmqGr3av7Mr8EngBAZ3U1Q0K1zKNm2NpYi2tXL8ba6iBSRKytDuLa1YutL0fty/wSeAIAndXVDAnVMo+aY2tjLW7tXI6Xd98Xt3Yutz7ojOjP/NJcCADoLOt7UgTziDL1ZX6lnHNlO9vc3Mz7+/uV7Q8AAIDqpJRu55w3z26X8QQAgIZp+rqOTR8fzSPwBACgdAKV+TV9Xcemj49mmtlcKKX07Smll1JKX0gpfT6l9NOj7d+WUvp0SukPR3++pfzhAgDQNieByvDwKHI8DFT2DoZ1D62Rmr6uY9PHRzPN09X26xHxwZzzuyPieyLiJ1NK746InYj4TM75nRHxmdG/AQDgFIHKYpq+rmPTx0czzQw8c86v5px/b/T3P4+IL0bEWkS8PyI+NnrZxyJiq6xBAgDQXgKVxTR9Xcemj49mWmgdz5TS0xGxERG/GxFvyzm/OvrSn0TE2wodGQAAnSBQWUzT13Vs+vhoprkDz5TSN0fEJyPiZ3LOf/bo1/Lxmixj12VJKT2XUtpPKe2/9tpr5xosAADtI1BZzNbGWly7ejHWVgeRImJtdRDXrl48V+OevYNhXNq9Gc/svBiXdm+e6/naMsZH9821jmdKaSUifisibuScf3G07W5EvCfn/GpK6e0R8U9yzlOvHtbxBADopyq62uqcO97ZLrQRx4G/YJEyLL2OZ0opRcSvRMQXT4LOkU9FxE9ExO7oz98saKwAAHTM1sZaqUGOJT4mm9bcqe/HhurMs47npYj48Yi4k1L67Gjbz8VxwPmJlNIHIuKPIuJHyhkiAFAH2SOKUNU8ElxNprkTTTAz8Mw5/9OISBO+/L3FDgcAaALZI4pQ5TwSXE321OoghmOOg+ZOVGmhrrYAQD9Yd5EiVDmPdM6dTHMnmmCeUlsAoGdkj1jUuJLaKufR9pX1sQ10BFcPs8tK56mTwBMAeIzSPBYxqaR29c0r8fob9x57fRnzSHA1XdnNnWAWgScA8BjZIxYxqaT2L73piRisXKhsHgmuoLkEngB0ms6sy5E9mp85Nrl09k+P7sWH/9Nne398AIEnAB2mM+v5yB7NZo4dm1aabR4BEbraAtBhOrNSNnPsmK6pwCwyngB0ls6s1etb2Wldc6xpx1lpNjCLwBOAztKZtVp9LDutY4419TgrqQWmUWoLQGcp/6tWH8tO65hjfTzOddo7GMal3ZvxzM6LcWn3ZuwdDHu1fyiKjCcAnaX8r1p9LG2uY4718TjXpe7sct37hyIJPAHoNOV/1elraXPVc6yvx7kO07LLVbznde8fiqTUFgAohNLmajjO1ak7u1z3/qFIAk8AoBBbG2tx7erFWFsdRIqItdVBXLt6UWamYI5zdSZlkavKLte9fyiSUlsAoDBKm6vhOFdj+8r6qWcsI6rNLte9fyiSwBMAKta0NRiB8epuUFb3/qFIKedc2c42Nzfz/v5+ZfsDgKY526Uy4jiDoVQSgC5IKd3OOW+e3S7jCQAV0qWSJpF9B6oi8ASACulSyaPqDPysEQlUSVdbAKiQLpWcOAn8hodHkeNh4Ld3MKxk/9Oy7wBFE3gCQIWswciJugM/2XegSgJPAKiQNRg5UXfgJ/sOVMkzngBQMWswEnEc4A3HBJlVBX7WiASqJPAEoHF02iyfY1y/ugO/staINLeAcQSeADSKTpvlc4yboazAb9ExFLk/cwuYJOWcK9vZ5uZm3t/fr2x/ALTPpd2bY8sP11YHcWvncg0j6h7HmLKYW0BK6XbOefPsds2FAGiUuhuu9IFjTFnMLWASgScAjaLTZvkcY8pibgGTCDwBaBTrXJbPMaYs5hYwieZCADRKExqudJ1jTFnMrWbQWZgm0lwIAAA64mxn4YjjrPO1qxcFn1RCcyEAAOi46zfungo6IyKO7t2P6zfu1jQiOKbUFgCAmZRvtkNTOgubL5wl8AQAYKqz5ZvDw6N4/oU7ERGCiYZ5anUwdi3VKjsLmy+Mo9QWAEq2dzCMS7s345mdF+PS7s3YOxjWPSQW1Pf3UPlmezShs7D5wjgyngC0XpNLutz5bz/vYXPKN5mtCZ2FzRfGEXgC0GpNDwqm3flvwviYzXvYjPJN5re1sVbr3DRfGEepLQCt1vSSLnf+59fUclbv4fjyzYiIN/7N1xvzPtEcTSj3pXlkPAFotaYHBe78z6fJmeu+vodnS9h/6LvW4rd+/9U4PLr34DWvv3GvMe8TzdGEcl+aR8YTgFab9OG/KUGBO//zaXLmuo/v4cmNgOHhUeQ4vhHwydvDSOnx1zblfaJZtjbW4tbO5Xh5931xa+eyoBMZTwDabfvK+qlMWUSzgoKm3vlvWkOmJmeum/oelmnSjYCz20404X0Cmk3gCUCrtSEoqLvRx1lNLGttejlr097Dsi0aSDblfQKaS+AJQOv1LSg4ryZ2aW165rpvJt0IWB2sxF98/RveJ2BhAk+AFmlaeSTt1MSy1jZkrvtk0o2AD/3H3xER3idgcQJPgJZoYnkk7dTUslaZ6+aYdSPA+wQsSuAJ0BJNLI+knZS11meZqoW6Kh3cCACKJPAEaIkmlkfSTspa67FM1YJKB6ArBJ4ALdHU8kjaSTarestULah0ALriiboHAMB8+riIPXTJMlULKh2ArhB4ArTE1sZaXLt6MdZWB5EiYm11ENeuXpT1gJaYVJ0wrWphme8BaCKltgAtojwS2muZpk4aQQFdIfAEAKjAMk2dNIICuiLlnCvb2ebmZt7f369sfwAAAOPUtVRR16WUbuecN89ul/EEAAB6xVJF1RN4ArSQu7QAsDxLFVVP4AnQMm26SytAhuZxXhbL8Wyn8y5V5H1fnOVUAFpm2l3aJjkJkIeHR5HjYYC8dzCse2jQW87LYjme7XWepYq878sReAK0TFsWlG9LgAx94rwsluPZXttX1mOwcuHUtnmXKvK+L0fgCdAybVlQvi0BMvSJ87JYjmd7bW2sxbWrF2NtdRApItZWB3Ht6sW5ymW978vxjCdAy7RlQfmnVgcxHPNLuGkBMvSJ87IYJ8/3TVqU0PFsh62NtaWey3QeLUfGE6BlznOXtkrnKWOCKu0dDOPS7s14ZufFuLR7s9PPaTkvz+/R5/vGcTy7z3m0HBlPgBZa9i5tlU7Gp+sfTdamLtFFcF6e37jn+06sOZ694DxaTsp5UpFA8TY3N/P+/n5l+wOAeWiL323T3t9LuzfHZq7WVgdxa+dy1UNlQXWcu8/svDi2xDZFxMu77yt139AGKaXbOefNs9tlPAHotb5lvPpm1vurSUh71XXuer4PluMZTwB6TVv8bpv1/ralSzSPq+vc9XwfLEfgCfRan5qKMJ6MV7fNen8FEe1V17nblgZv0DRKbYHeUmJ5fl14NlLZXLfNen/P2ySkC+dAW9V57rahwRs0jYwn0FtKLM/n0SUFcjwM3NuWNZbx6rZ53t+tjbW4tXM5Xt59X9zaubxQ0NmFc6CtnLvQLgJPoLeUWJ5PVwJ3ZXPdVub725VzoK2cu9AuM0ttU0ofjYi/HRFfyzn/tdG2D0XEfxERr41e9nM5598ua5AAZVBieT5dCtzHlc0poeyOssoiu3QOtJWSV6rg90Ex5sl4/mpEfP+Y7R/OOT87+k/QCbSOMq3z6XI3UCWUzKPL5wBwzO+D4swMPHPOvxMR/6qCsQBUSpnW+XQ5cFdCWb4mdpRedExdPgeAY34fFOc8XW1/KqX0dyJiPyI+mHN+fdyLUkrPRcRzERHveMc7zrE7gOIp01reebuBNpkSynI1saP0MmPq8jkAHPP7oDjLBp6/FBH/dUTk0Z+/EBH/+bgX5pw/EhEfiYjY3NzMS+4PgAbqauDu+d9yTcsg1DWflh1TV8+BJvO8HUWYdx75fVCcpbra5py/mnO+n3P+RkT8ckR8d7HDApqqieVxUDQllOVqWgZh72A49oNlhKxG03jejiIsMo/8PijOUoFnSuntj/zzByPic8UMB2gyv/DpC8//lqtJTXlOrmuTyGo0i+ftKMIi88jvg+LMs5zKr0XEeyLirSmlP46In4+I96SUno3jUtuvRMTfLXGMQEPMW4qmDIouWLSE0ryf3/aV9VPPU0bUl0EYd107IavRPE3LltNOi84jJfXFmBl45px/bMzmXylhLEDDzXOhbmLTEChb2+d91UFzk5ryTAtYZDWax/N2FME8qsdSpbZAP81THqcMij5q87yvq4R+a2Mtbu1cjpd33xe3di7XFuBNuq6trQ4EnQ3keTuKYB7VQ+AJzG2eC7UyKPqozfO+zKC5Dc3IfABtF8/bUQTzqB7nWccT6Jl5yuOUr9BHbZ73ZQXNbSk/blLZL/PxvB1FMI+qJ/AEFjLrQt2kpiFQlTbP+7KC5iau1TmJD6AA5RN4AoWSPege3Vpna/O8LytobnP5MQDFE3gChZM96I62lEs2QVvnfVlB87yZVDc2APpB4AnARG0ql2R5ZQTN82RS3dgA6A+BJ0DJ2pzRUS7JsubJpLqxAdAfAk+AErU9o9Pmbq3Ub1Ym1Y0NgP6wjidAicpcI7EK1jikTJNuYLixAdA9Ak+AErU9o2ORbcrkxgZAfyi1BShRF0pV29qtleZr8zI0ACxG4AlQorLWSISucGMDoB8EntATbe6s2mYyOgAAAk/ohbZ3Vm07GR0AoO80F4IeaHtnVQAA2k3GE3qg7Z1VoUjKzqFdijpnnftQL4En9EAXOqtCEZSdQ7sUdc4696F+Sm2hB7qwVt7ewTAu7d6MZ3ZejEu7N2PvYFj3kGghZefQLkWds859qJ+MJ/RA2zurulNdrzaVp80aq7JzaJeiztkyz/02XSOhTgJP6Ik2d1addqe6rT9TW7Qp6J9nrMrOoV2KOmfLOvfbdI2Euim1BRpPlqo+bSpPm2esXSg7hz4p6pwt69xv0zUS6ibjCTSeLFV92hT0zzPWtpedQ98Udc6Wde636RoJdRN4Ao23fWX9VClThCxVVdoU9M871jaXnUMfFXXOlnHut+kaCXVTags03tbGWly7ejHWVgeRImJtdRDXrl588AFCx9vy1F2aush7W/dYqZbzfjrHpxquOzC/lHOubGebm5t5f3+/sv0B3Xe2sUPE8S/9RwNTzqeujo3LvLe6S/aD8346x6darjtwWkrpds5587HtAk+gzS7t3hxb5rS2OohbO5drGBFF8d72z7wf4M2N6RwfaKa+3KSYFHh6xhNoNY0dust72y+LLEthbkzn+FC3vgRYi7D0jmc8gZab1MBBY4f28972yyLLUpgb0zk+1OkkwBoeHkWOhwFW358ztvSOwBNYQpOaVmjs0F3e23rUdX4vkqUzN6ZzfKiTAGs8lQhKbaEXiix5aVqpiHUZu8t7W706z+9FlqUwN6ZzfIqlbHQxAqzxLL2juRB0XtHdDTWtgO6q8/zWiZUmMi8X53PCeH2aS5OaCym1hY4ruuTFncx6NanMme6p8/yetV4v1EHZ6OKUeo/nGqfUFjqv6A+SSkXq07QyZ7qn7vN7a2PNXKZRZv0OVYb7OKXek/X9GifjCR1XdHdDdzLr4847ZXN+w2nTfofq3jrZ1sZa3Nq5HC/vvi9u7VzudbDFQwJP6LiiP0gqFamPMudiKFeezPndDXXP8br3X6Rpv0PdDITFKLWFjiuj5KXvpSJ1qbsMsguUK8/m/G63uud43fsv2rTfoT/78c+O/R43A2E8gSf0gA+S3bB9ZX1sRzxlkPOblqFwjtAFdc/xuvdfhkm/Q90MhMUIPIFz0VihOho2nJ9yZbqu7jle9/6r1PebgX7/syiBJ7C0rpVUtYHs9fnIUNB1dc/xuvdfpT7fDPT7n2UIPIGldbGkqu0WvQPdtzvWfc9Q0H11z/G697+M81wH+3oz0O9/liHwhIp16YN+n0qq2mDRO9B9vGPd5wwF/VD3HK97/4vq43WwCE3+/d+lz1ldI/CECnXtF9zqm1fi9Tfujd1O9Ra9A93XO9Z9zVDQH3XP8br3v4i+XgfPq6kl1V37nNU1Ak+oUNd+weW82HbKtegd6Cbfsaa5xmUTItqT4aqbbMxpdR8P18HlNLWkumufs7pG4AmPKPsXYNd+wf3p0ePZzmnbKdeid6AXfX3dHxCp37hswvb/9vsRKeLe/fxgmwzDeLIxp02aT3/vf/98HL5xr5LrTFMzd03X1JLqrn3O6pon6h4ANMXJL8Dh4VHkePiBYO9gWNg+pgUAbdS1n6fttq+sx2Dlwqlt0+5AL/L6Ks4Pmm9cNuHeN/KDoPPESYaB06ZlY/po0nx6/Y17lV1nFr1u8tDWxlrc2rkcL+++L27tXK496IzwuaTpBJ4wUsUHgq79guvaz9MUewfDuLR7M57ZeTEu7d6c+0PX1sZaXLt6MdZWB5EiYm11ENeuXpz4YWCR1/vATMRiWQMZhsfJxpw2z89d9nVm0esmzeZzSbMptYWRKj4QNLU0ZVld+3nOqqO09LyleIs29Zj39T4wEzG5LHHSazltWllnH0vZv3WwEodzPJpR9nWmTc2QmK7rn0vaTuAJI1U959G1X3Bd+3lO1PUsVlMbI3gOiojxDUVWnkinnvGMkGGYZFJDlve+68lePvuZ0nyvc51hEV39XNIFSm1hpE/lGcuWchb1/W1QV2lpUzOLfTo/mGxcWeL1/+Q74/oPf6dSxTlMKut86Uuv9bKU/XDMclxnuc5Ad8h4wkhfyjPOm8nrS1fGugLApmYW+3J+MNukbIK5MJ9xx+9nP/7Zsa+t+4ZT2SZd7y6kFN/I2XUGOkbgCY/oQ3nGeUs5m1oKWrS6AsCmro0W0bzzo4/PxNFNTb3hVLZJ1zsZc+gmpbbQM+fN5DW1FLRodZWW6rA4H8u70CV9LWV3vYN+kfGEnjnvnfW+3Jmvs7S0aZnFJupL5p1+6HMpu+sd9IfAE3rmvKWcTS4FLVpXPxA1sUR10TH1JfNOf3T1egNwQuBJozXxA3LbLXpnfdx7cO3qRe9LSzWxOdQyY+pL5h0AuiLlnGe/qiCbm5t5f3+/sv3Rbmc/jEZoOlA170H3XNq9OTZgW1sdxK2dyzWMaLkxmZsA0Ewppds5582z22U8aZyTDNu4D6JH9+7HBz/x+/GzH/+sTFsFPEfXPU0sUV1mTH1+Jg4A2kjgSaOMy2KcdX+UpW9CiWDXNTFIaaq2lIU3sUR12TF5Jg4A2sNyKjTKuAzbNCfZN8ox6YO/5+hOa9PSHk1ctqGJYwIAiiXjSaMsk0mTfStPkzvYTsswVp19bFNJchNLVJs4JqhaW6omAJYl8KRRJpXcRURcSOlBme3Z76EcTQ0IpnVBjYjKu7a2rSS5iSWqTRwTVKWJ3aYBiibwpFEmZdiuXb0YEdHY7FubzbrL3sSAYFqG8eTv475W1s/RxOcm6SZZsW5qU9UEwM5DQAUAABPlSURBVLIEnjTKPBm2ZT50+bA2Xlvvsi+TYSwz+7hMSbI5yaLaer4yW9uqJgCWIfCkcaZl2JbJvvmwNllb77LPyjBWnX1ctCTZnGQZbT1fmU3VBNAHAk86z4e1ydp6l31WhrGOkuxFboqYk/V7dL3gk+fH1xqeeW7r+cpsTW7kBlAUgSed58PaZG29y15WSXZVzMl6nc04t2Vt4Laer8zW1EZuAEUSeNJ5PqxN1ua77EWXZFfJnKzXtPWCm5x5bvP5ymxNv24BnJfAk056tHHL6ptX4omI+MYjX195IvmwFu6y10UAUa9ZmeUyMs9FNJNyvgLQZjMDz5TSRyPib0fE13LOf2207dsi4uMR8XREfCUifiTn/Hp5w4T5nS2je/2Ne4+/KFU8qAZzl716Aoh6TVsv+OTrRSqymZTzFYC2emKO1/xqRHz/mW07EfGZnPM7I+Izo39DI0wroztx735+sOYj1GFrYy1u7VyOl3ffF7d2LgsmKrR9ZT0GKxfGfq2MzPOsdWcBoA9mZjxzzr+TUnr6zOb3R8R7Rn//WET8k4j4LwscFyxt3jI5jVzOxzqUtNWjGecqutpqJgUAyz/j+bac86ujv/9JRLxt0gtTSs9FxHMREe94xzuW3B3Mb1YZ3aOvYznWoaTtqixZ1UwKAOYrtZ0q55wjIk/5+kdyzps5580nn3zyvLuDmaaV0Z3QyOV8lA7C/MZdk1yDAOibZTOeX00pvT3n/GpK6e0R8bUiBwXnMa5xy3vf9WS89KXXGl0WOqt0ddHS1jJLYbtWOjjPsVJazLI0kwKA5QPPT0XET0TE7ujP3yxsRFCAtnV+nFW6umhpa9mlsF0qHZznWCkt5rzadk0CgKLNLLVNKf1aRPyfEbGeUvrjlNIH4jjg/JsppT+MiP9w9G9ojL2DYVzavRnP7LwYl3Zvxt7BsO4hTTWrdHXR0tayS2G7VDo4z7H60Kc+r7QYAOAc5ulq+2MTvvS9BY+FBjlvWWGdZYltzE7NKl1dtLS17FLYLpUOTjomw8OjeGbnxVh980ocHo1ZC3bK9wIAcNqypbZ02HkDt7oDv2kZrKYGRrNKVxctba2iFLYrpYPTuiDniHj9jfFB58n3AgAw27m72tI95y3TrLvjaRsb38wqXV20tLXKUti2lTWfNU8X5GnfCwDAbDKeLVZWOet5A7e6A782Nr6ZVbq6aGnrtNcXOW/qzm4X4eyxmrg21BlvefNKa35GAIC6CTxbqswP/OcN3OoO/LavrJ86NhHtaHwzq3R10dLWca8vet60sax5nEeP1aXdmxNLb08MVi7Ez/9H31HF0AAAOkGpbUsVWc56tlTyve968lxlmnV3PN3aWItrVy/G2uogUkSsrQ7i2tWLrQqEylJ0GXTd2e0yjJu/KxdSrA5WzCcAgCXJeLZUUR/4x2XAPnl7GD/0XWvx0pdeW6ocswkdT7vS+KZoRQeKdWe3y9CE+QsA0DUCz5Yq6gP/pAzYS196LW7tXF56fAK/Zio6UGxrWfMs5i8AQLGU2rZUUeWsXSyVZLKiy6CVNQMAMA8Zz5Yqqhywi6WSTFZGGansIAAAs6Sc51084Pw2Nzfz/v5+ZftjtrPPeEYcZ8BkrQAAgEWllG7nnDfPbpfx7Ll5MmBlrRdK/5hLAAD9JPBkaqlkmeuF0i/mEgBAfwk8W6qqzNG0dR8FC/OT6atnLjnuAADNIPBsoSozR7renl/Vmb6mBltVzyUZVgCA5rCcSgtNyxwVbVJ3W11v51fl+3USbA0PjyLHw2Br72BY+L4WVfVcqvK4AwAwncCzhcrOHO0dDOPS7s14ZufF+P/+4uuxciGd+vp51n3soyozfU0OtopeQ3QW2XoAgOZQattCZa69ebY88fDoXqw8keItb16JwzfuNap0syrnLV2tcq3UJgdbZawhOo01agEAmkPg2ULbV9bHrr1ZROZoXMbs3jdyvPmb3hQH/9X3nfv/3zZFPCdY5vt1VtODrWkdlItW5XEHAGA6pbYttLWxFteuXoy11UGkiFhbHcS1qxcL+UDf5IxZHYooXS3z/Tqr6nLWJqvyuAMAMJ2MZ0uVlTlqesasakUF4lVl+qouZ226KjOsAABMJvDkFOWJp7UxEBdsAQDQNEptOUV54mnzlq4+2gn40u7NRixfAgAATSHj2QHn7bp6lozZQ/OUrhbRgAgAALpM4Nlygp7yzQrEpzUg8h4AAIBS29Yrousq56MTMAAATCfwbDlBT/0mNRpqcgMiAACoksCz5QQ99bN2JgAATCfwbDlBT/10AgYAgOk0F2q5ebquUj6dgAEAYDKBZwecDXpO1pTsQiBa9FIxAABA9QSeHdOl5VW69LMAAECfecazY7q0vEqXfhYAAOgzgWfHdGl5lS79LAAA0GdKbWtW9DOMT60OYjgmMCt6eZUqnr2s6mcBAADKJeNZo5NnGIeHR5Hj4TOMewfDpf+fVSyvUsa4x7FUDAAAdIOMZ4lmZQWnPcO4bPbwvMurzJPJLGPc41gqBgAAukHgWZJ5OrKW9QzjsmtKzttFtspnL62PCQAA7afUtiTzdGSd9KxiXc8wzttFtmnjBgAAmk3gWZJ5soJFP8O4dzCMS7s345mdF+PS7s2Fn7mcN5Pp2UsAAGARSm1LMk9H1iKfYZy3TPa8Yy563AAAQPcJPEuyfWX9VCAYMT4rWNQzjEU0/Jk15iqWUAEAALpH4FmSqrOCRTT8mTbmIjKqAABAPwk8S1RlR9Z5y2RnmTTmqpZQAQAAukdzoY7YvrIeK0+kU9tWnkiFNfypcgkVAACgW2Q8S1Dbs5Cn4864940cH/rU5yPi/OWwRWVUAQCA/pHxLNjJs5DDw6PI8fBZyEWXNlnU9Rt34979/Nj2w6N7hezfEioAAMCyBJ4Fm/YsZJmmlbwWsf+tjbW4dvVirK0OIkXE2uogrl296PlOAABgJqW2BavrWchJpbBF7r/KZkkAAEB3yHgWbNIzj2U/CzmuFLbK/QMAAEwi8CxYXc9CnpTCvuXNK499zbOYAABAnZTaFuykFLWOrrYnpbC1ddUFAAAYI+X8eCfUsmxubub9/f3K9gcAAEB1Ukq3c86bZ7crtQUAAKBUAk8AAABKJfAEAACgVJoLtZgmQgAAQBsIPFtq72AYz79wJ47u3Y+IiOHhUTz/wp2ICMEnAADQKEptW+r6jbsPgs4TR/fux/Ubd2saEQAAwHgyniUqsxT2lcOjhbYDAADURcazJCelsMPDo8jxsBR272BYyP//qdXBQtsBAADqIvAsybKlsHsHw7i0ezOe2XkxLu3enBiobl9Zj8HKhVPbBisXYvvK+vkGDgAAUDCltgU7Ka8dLlEKu0jDoJN/62oLAAA0ncCzQGcDx3GmlcJOy5KOCyi3NtYEmgAAQOMptS3QuMDxUbNKYTUMAgAAukjGc0njOtZOCxDX5iiFfWp1MLZEV8MgAACgzWQ8lzCpY+23DlbGvn5tdRC3di7PLIt977ueXGg7AABAGwg8lzDpWcyU4lydZl/60msLbQcAAGgDgecSJpXUHr5xL65dvRhrq4NIcZzpvHb14twNgDzjCQAAdNG5nvFMKX0lIv48Iu5HxNdzzptFDKrppj2LeZ5Os57xBAAAuqiIjOd7c87P9iXojIjYvrJ+rpLaqv+/AAAAddLVdgknGc2zXW3Pu6ZmWf9fAACAOqWc8/LfnNLLEfF6ROSI+F9yzh+Z9vrNzc28v7+/9P4AAABorpTS7XHVsOfNeP6NnPMwpfTvRMSnU0pfyjn/zpkdPxcRz0VEvOMd7zjn7tpl3FqfspcAAEDfnOsZz5zzcPTn1yLiNyLiu8e85iM5582c8+aTT/ZnPcpJa33uHQzrHhoAAECllg48U0p/OaX0LSd/j4jvi4jPFTWwtpu01uf1G3drGhEAAEA9zlNq+7aI+I2U0sn/5x/mnP9xIaPqAGtyAgAAHFs68Mw5/4uI+M4Cx9Ip1uQEAAA4VsQ6noxhTU4AAIBj1vEsiTU5AQAAjgk8S7S1sSbQBAAAek+pLQAAAKUSeAIAAFAqgScAAAClEngCAABQKoEnAAAApRJ4AgAAUCqBJwAAAKUSeAIAAFAqgScAAAClEngCAABQKoEnAAAApRJ4AgAAUCqBJwAAAKUSeAIAAFAqgScAAAClEngCAABQKoEnAAAApXpT3QNok72DYVy/cTdeOTyKp1YHsX1lPbY21uoeFgAAQKMJPOe0dzCM51+4E0f37kdExPDwKJ5/4U5EhOATAABgCqW2c7p+4+6DoPPE0b37cf3G3ZpGBAAA0A4Czzm9cni00HYAAACOCTzn9NTqYKHtAAAAHBN4zmn7ynoMVi6c2jZYuRDbV9ZrGhEAAEA7aC40p5MGQrraAgAALEbguYCtjTWBJgAAwIKU2gIAAFAqgScAAAClEngCAABQKoEnAAAApRJ4AgAAUCqBJwAAAKUSeAIAAFAqgScAAAClEngCAABQKoEnAAAApRJ4AgAAUCqBJwAAAKUSeAIAAFAqgScAAAClEngCAABQKoEnAAAApRJ4AgAAUKo31T2Aptg7GMb1G3fjlcOjeGp1ENtX1mNrY63uYQEAALSewDOOg87nX7gTR/fuR0TE8PAonn/hTkSE4BMAAOCclNpGxPUbdx8EnSeO7t2P6zfu1jQiAACA7hB4RsQrh0cLbQcAAGB+As+IeGp1sNB2AAAA5ifwjIjtK+sxWLlwattg5UJsX1mvaUQAAADdoblQPGwgpKstAABA8QSeI1sbawJNAACAEii1BQAAoFQCTwAAAEol8AQAAKBUAk8AAABKJfAEAACgVAJPAAAASiXwBAAAoFQCTwAAAEol8AQAAKBUAk8AAABKJfAEAACgVAJPAAAASiXwBAAAoFQCTwAAAEol8AQAAKBUKedc3c5Sei0i/qiyHS7mrRHxL+seBI1mjjCLOcIs5gjzME+YxRxhljrnyL+bc37y7MZKA88mSynt55w36x4HzWWOMIs5wizmCPMwT5jFHGGWJs4RpbYAAACUSuAJAABAqQSeD32k7gHQeOYIs5gjzGKOMA/zhFnMEWZp3BzxjCcAAAClkvEEAACgVL0PPFNK359SuptS+nJKaafu8dAMKaWvpJTupJQ+m1LaH237tpTSp1NKfzj68y11j5NqpZQ+mlL6Wkrpc49sGzsv0rH/YXRt+YOU0l+vb+RUZcIc+VBKaTi6nnw2pfQDj3zt+dEcuZtSulLPqKlSSunbU0ovpZS+kFL6fErpp0fbXUuIiKlzxLWEiIhIKf1bKaV/llL6/dEc+Xuj7c+klH53NBc+nlL6ptH2vzT695dHX3+6jnH3OvBMKV2IiP8pIv5WRLw7In4spfTuekdFg7w35/zsI62odyLiMznnd0bEZ0b/pl9+NSK+/8y2SfPib0XEO0f/PRcRv1TRGKnXr8bjcyQi4sOj68mzOeffjogY/b750Yj4jtH3/M+j30t029cj4oM553dHxPdExE+O5oJrCScmzZEI1xKO/UVEXM45f2dEPBsR359S+p6I+G/jeI78exHxekR8YPT6D0TE66PtHx69rnK9Djwj4rsj4ss553+Rc/43EfHrEfH+msdEc70/Ij42+vvHImKrxrFQg5zz70TEvzqzedK8eH9E/K/52P8VEasppbdXM1LqMmGOTPL+iPj1nPNf5Jxfjogvx/HvJTos5/xqzvn3Rn//84j4YkSshWsJI1PmyCSuJT0zuh7869E/V0b/5Yi4HBH/aLT97HXk5PryjyLie1NKqaLhPtD3wHMtIv6fR/79xzH9xKY/ckT8Hyml2yml50bb3pZzfnX09z+JiLfVMzQaZtK8cH3hUT81KpP86CNl+uZIz43K3TYi4nfDtYQxzsyRCNcSRlJKF1JKn42Ir0XEpyPi/46Iw5zz10cveXQePJgjo6//aUT829WOWOAJk/yNnPNfj+MSp59MKf0Hj34xH7eD1hKaU8wLJviliPircVwO9WpE/EK9w6EJUkrfHBGfjIifyTn/2aNfcy0hYuwccS3hgZzz/ZzzsxHxV+I4w/2umoc0U98Dz2FEfPsj//4ro230XM55OPrzaxHxG3F8Qn/1pLxp9OfX6hshDTJpXri+EBEROeevjj4gfCMifjkelsCZIz2VUlqJ44DiH+ScXxhtdi3hgXFzxLWEcXLOhxHxUkT8+3Fciv+m0ZcenQcP5sjo698aEf9vxUPtfeD5zyPinaMOUN8Uxw9mf6rmMVGzlNJfTil9y8nfI+L7IuJzcTw3fmL0sp+IiN+sZ4Q0zKR58amI+DujjpTfExF/+kgZHT1y5nm8H4zj60nE8Rz50VG3wWfiuHnMP6t6fFRr9FzVr0TEF3POv/jIl1xLiIjJc8S1hBMppSdTSqujvw8i4m/G8bPAL0XED49edvY6cnJ9+eGIuDmqrKjUm2a/pLtyzl9PKf1URNyIiAsR8dGc8+drHhb1e1tE/Mbomes3RcQ/zDn/45TSP4+IT6SUPhARfxQRP1LjGKlBSunXIuI9EfHWlNIfR8TPR8RujJ8Xvx0RPxDHTR7eiIj/rPIBU7kJc+Q9KaVn47h08isR8XcjInLOn08pfSIivhDHXSx/Mud8v45xU6lLEfHjEXFn9HxWRMTPhWsJD02aIz/mWsLI2yPiY6PuxU9ExCdyzr+VUvpCRPx6Sum/iYiDOL6BEaM//35K6ctx3ADvR+sYdKoh2AUAAKBH+l5qCwAAQMkEngAAAJRK4AkAAECpBJ4AAACUSuAJAABAqQSeAAAAlErgCQAAQKkEngAAAJTq/we0spVn+rZS6AAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 1152x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tIkUgpzxhP9n",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model = tf.keras.Sequential()\n",
        "model.add(tf.keras.layers.Dense(1, input_shape=[1]))\n",
        "model.compile(loss='mean_squared_error', optimizer=tf.keras.optimizers.Adam(0.01))"
      ],
      "execution_count": 123,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mP8wk96ahVIa",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 191
        },
        "outputId": "bc451af0-8bfa-411b-ebea-fd79846cb8fe"
      },
      "source": [
        "model.summary()"
      ],
      "execution_count": 124,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_11\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "dense_14 (Dense)             (None, 1)                 2         \n",
            "=================================================================\n",
            "Total params: 2\n",
            "Trainable params: 2\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Kad2kh20hapA",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "d380d82a-dba6-4d43-9f9d-e471c3662a8e"
      },
      "source": [
        "history = model.fit(df[\"X\"], df[\"y\"], epochs=300)"
      ],
      "execution_count": 125,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 12909.1787\n",
            "Epoch 2/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 10396.9492\n",
            "Epoch 3/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 8191.7358\n",
            "Epoch 4/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 6346.2607\n",
            "Epoch 5/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 4751.0576\n",
            "Epoch 6/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 3500.0896\n",
            "Epoch 7/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 2500.7258\n",
            "Epoch 8/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 1752.5309\n",
            "Epoch 9/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 1182.4645\n",
            "Epoch 10/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 787.9683\n",
            "Epoch 11/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 511.8502\n",
            "Epoch 12/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 320.3270\n",
            "Epoch 13/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 200.8797\n",
            "Epoch 14/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 122.4936\n",
            "Epoch 15/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 78.8654\n",
            "Epoch 16/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 51.2773\n",
            "Epoch 17/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 38.3485\n",
            "Epoch 18/300\n",
            "7/7 [==============================] - 0s 4ms/step - loss: 31.2884\n",
            "Epoch 19/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 27.8213\n",
            "Epoch 20/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 26.1903\n",
            "Epoch 21/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.7626\n",
            "Epoch 22/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.5160\n",
            "Epoch 23/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 25.5411\n",
            "Epoch 24/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.4754\n",
            "Epoch 25/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.4779\n",
            "Epoch 26/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.4632\n",
            "Epoch 27/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.4423\n",
            "Epoch 28/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 25.4464\n",
            "Epoch 29/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 25.4149\n",
            "Epoch 30/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.3906\n",
            "Epoch 31/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 25.3817\n",
            "Epoch 32/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.3751\n",
            "Epoch 33/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.3614\n",
            "Epoch 34/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.3502\n",
            "Epoch 35/300\n",
            "7/7 [==============================] - 0s 4ms/step - loss: 25.3162\n",
            "Epoch 36/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.3068\n",
            "Epoch 37/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.2861\n",
            "Epoch 38/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.2657\n",
            "Epoch 39/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 25.2481\n",
            "Epoch 40/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.2280\n",
            "Epoch 41/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.2157\n",
            "Epoch 42/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.1860\n",
            "Epoch 43/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.1675\n",
            "Epoch 44/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.1501\n",
            "Epoch 45/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 25.1325\n",
            "Epoch 46/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.1342\n",
            "Epoch 47/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.1075\n",
            "Epoch 48/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.1083\n",
            "Epoch 49/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.0835\n",
            "Epoch 50/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.0284\n",
            "Epoch 51/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 25.0301\n",
            "Epoch 52/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 25.0077\n",
            "Epoch 53/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.9676\n",
            "Epoch 54/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.9611\n",
            "Epoch 55/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.9558\n",
            "Epoch 56/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.9279\n",
            "Epoch 57/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.8868\n",
            "Epoch 58/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.8594\n",
            "Epoch 59/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.8317\n",
            "Epoch 60/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.8374\n",
            "Epoch 61/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.8076\n",
            "Epoch 62/300\n",
            "7/7 [==============================] - 0s 4ms/step - loss: 24.7862\n",
            "Epoch 63/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.7552\n",
            "Epoch 64/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 24.7225\n",
            "Epoch 65/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.7156\n",
            "Epoch 66/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.6919\n",
            "Epoch 67/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 24.7085\n",
            "Epoch 68/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.6413\n",
            "Epoch 69/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.6879\n",
            "Epoch 70/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.6034\n",
            "Epoch 71/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.5884\n",
            "Epoch 72/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.5675\n",
            "Epoch 73/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.5292\n",
            "Epoch 74/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.5105\n",
            "Epoch 75/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.4944\n",
            "Epoch 76/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.4624\n",
            "Epoch 77/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.4200\n",
            "Epoch 78/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 24.4106\n",
            "Epoch 79/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.3647\n",
            "Epoch 80/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 24.3516\n",
            "Epoch 81/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.3234\n",
            "Epoch 82/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 24.3468\n",
            "Epoch 83/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.2740\n",
            "Epoch 84/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.2682\n",
            "Epoch 85/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.2219\n",
            "Epoch 86/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.1728\n",
            "Epoch 87/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.1493\n",
            "Epoch 88/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 24.1372\n",
            "Epoch 89/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.1042\n",
            "Epoch 90/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.0807\n",
            "Epoch 91/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.0461\n",
            "Epoch 92/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.0029\n",
            "Epoch 93/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 24.0199\n",
            "Epoch 94/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.9650\n",
            "Epoch 95/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.9714\n",
            "Epoch 96/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.9086\n",
            "Epoch 97/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.9103\n",
            "Epoch 98/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.8840\n",
            "Epoch 99/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.8398\n",
            "Epoch 100/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.7944\n",
            "Epoch 101/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 23.7823\n",
            "Epoch 102/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.7979\n",
            "Epoch 103/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.7587\n",
            "Epoch 104/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.6996\n",
            "Epoch 105/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.6504\n",
            "Epoch 106/300\n",
            "7/7 [==============================] - 0s 4ms/step - loss: 23.6388\n",
            "Epoch 107/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.6203\n",
            "Epoch 108/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.5875\n",
            "Epoch 109/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.5443\n",
            "Epoch 110/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.5139\n",
            "Epoch 111/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.4725\n",
            "Epoch 112/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.4533\n",
            "Epoch 113/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.4282\n",
            "Epoch 114/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.4232\n",
            "Epoch 115/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.3833\n",
            "Epoch 116/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.3285\n",
            "Epoch 117/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.3056\n",
            "Epoch 118/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 23.2734\n",
            "Epoch 119/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.2227\n",
            "Epoch 120/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.1977\n",
            "Epoch 121/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.1757\n",
            "Epoch 122/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.1398\n",
            "Epoch 123/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.1119\n",
            "Epoch 124/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.0837\n",
            "Epoch 125/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.0391\n",
            "Epoch 126/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 23.0178\n",
            "Epoch 127/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 22.9992\n",
            "Epoch 128/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 22.9348\n",
            "Epoch 129/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 22.9558\n",
            "Epoch 130/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 22.8900\n",
            "Epoch 131/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 22.8741\n",
            "Epoch 132/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 22.8194\n",
            "Epoch 133/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 22.7910\n",
            "Epoch 134/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 22.7485\n",
            "Epoch 135/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 22.6935\n",
            "Epoch 136/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 22.7285\n",
            "Epoch 137/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 22.6631\n",
            "Epoch 138/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 22.6527\n",
            "Epoch 139/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 22.6173\n",
            "Epoch 140/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 22.5559\n",
            "Epoch 141/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 22.5249\n",
            "Epoch 142/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 22.5392\n",
            "Epoch 143/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 22.4684\n",
            "Epoch 144/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 22.4248\n",
            "Epoch 145/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 22.3749\n",
            "Epoch 146/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 22.3371\n",
            "Epoch 147/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 22.3321\n",
            "Epoch 148/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 22.2845\n",
            "Epoch 149/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 22.2462\n",
            "Epoch 150/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 22.2088\n",
            "Epoch 151/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 22.1783\n",
            "Epoch 152/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 22.1504\n",
            "Epoch 153/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 22.1470\n",
            "Epoch 154/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 22.0974\n",
            "Epoch 155/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 22.0344\n",
            "Epoch 156/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 22.0082\n",
            "Epoch 157/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 21.9737\n",
            "Epoch 158/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 21.9474\n",
            "Epoch 159/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 21.9285\n",
            "Epoch 160/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 21.8645\n",
            "Epoch 161/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 21.8374\n",
            "Epoch 162/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 21.8093\n",
            "Epoch 163/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 21.7706\n",
            "Epoch 164/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 21.7592\n",
            "Epoch 165/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 21.6903\n",
            "Epoch 166/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 21.6766\n",
            "Epoch 167/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 21.6241\n",
            "Epoch 168/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 21.5972\n",
            "Epoch 169/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 21.5911\n",
            "Epoch 170/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 21.5939\n",
            "Epoch 171/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 21.5172\n",
            "Epoch 172/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 21.4584\n",
            "Epoch 173/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 21.4134\n",
            "Epoch 174/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 21.3625\n",
            "Epoch 175/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 21.3337\n",
            "Epoch 176/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 21.3107\n",
            "Epoch 177/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 21.2563\n",
            "Epoch 178/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 21.2244\n",
            "Epoch 179/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 21.1815\n",
            "Epoch 180/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 21.2415\n",
            "Epoch 181/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 21.1546\n",
            "Epoch 182/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 21.0611\n",
            "Epoch 183/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 21.0853\n",
            "Epoch 184/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 21.0374\n",
            "Epoch 185/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 21.0217\n",
            "Epoch 186/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.8966\n",
            "Epoch 187/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.9423\n",
            "Epoch 188/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 21.0291\n",
            "Epoch 189/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.9079\n",
            "Epoch 190/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.7599\n",
            "Epoch 191/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.7841\n",
            "Epoch 192/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.7472\n",
            "Epoch 193/300\n",
            "7/7 [==============================] - 0s 4ms/step - loss: 20.7045\n",
            "Epoch 194/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.6584\n",
            "Epoch 195/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.7054\n",
            "Epoch 196/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.6568\n",
            "Epoch 197/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.5370\n",
            "Epoch 198/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.5161\n",
            "Epoch 199/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.4530\n",
            "Epoch 200/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 20.3848\n",
            "Epoch 201/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 20.4104\n",
            "Epoch 202/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.3697\n",
            "Epoch 203/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.3545\n",
            "Epoch 204/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.3477\n",
            "Epoch 205/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.3362\n",
            "Epoch 206/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.1667\n",
            "Epoch 207/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.1794\n",
            "Epoch 208/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.1579\n",
            "Epoch 209/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.1085\n",
            "Epoch 210/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.1242\n",
            "Epoch 211/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 20.0176\n",
            "Epoch 212/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.9845\n",
            "Epoch 213/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 19.9487\n",
            "Epoch 214/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 19.8906\n",
            "Epoch 215/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.8585\n",
            "Epoch 216/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.8175\n",
            "Epoch 217/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 19.8667\n",
            "Epoch 218/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.7502\n",
            "Epoch 219/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.7219\n",
            "Epoch 220/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.6948\n",
            "Epoch 221/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.7738\n",
            "Epoch 222/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.5665\n",
            "Epoch 223/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.5572\n",
            "Epoch 224/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.5431\n",
            "Epoch 225/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.5357\n",
            "Epoch 226/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.4473\n",
            "Epoch 227/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.4026\n",
            "Epoch 228/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.6208\n",
            "Epoch 229/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.5252\n",
            "Epoch 230/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.2751\n",
            "Epoch 231/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.2713\n",
            "Epoch 232/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.2511\n",
            "Epoch 233/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.1729\n",
            "Epoch 234/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.2498\n",
            "Epoch 235/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.1664\n",
            "Epoch 236/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 19.0840\n",
            "Epoch 237/300\n",
            "7/7 [==============================] - 0s 4ms/step - loss: 19.0627\n",
            "Epoch 238/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.0197\n",
            "Epoch 239/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 19.0818\n",
            "Epoch 240/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 18.9180\n",
            "Epoch 241/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 18.9300\n",
            "Epoch 242/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 18.9398\n",
            "Epoch 243/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 18.8813\n",
            "Epoch 244/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 18.7979\n",
            "Epoch 245/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 18.7834\n",
            "Epoch 246/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 18.7463\n",
            "Epoch 247/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 18.6629\n",
            "Epoch 248/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 18.6386\n",
            "Epoch 249/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 18.6097\n",
            "Epoch 250/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 18.7732\n",
            "Epoch 251/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 18.5978\n",
            "Epoch 252/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 18.5262\n",
            "Epoch 253/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 18.5192\n",
            "Epoch 254/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 18.4141\n",
            "Epoch 255/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 18.4162\n",
            "Epoch 256/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 18.3787\n",
            "Epoch 257/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 18.3286\n",
            "Epoch 258/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 18.3045\n",
            "Epoch 259/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 18.2878\n",
            "Epoch 260/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 18.2149\n",
            "Epoch 261/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 18.1533\n",
            "Epoch 262/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 18.1875\n",
            "Epoch 263/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 18.0915\n",
            "Epoch 264/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 18.1969\n",
            "Epoch 265/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 18.1131\n",
            "Epoch 266/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.9771\n",
            "Epoch 267/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.9305\n",
            "Epoch 268/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.8862\n",
            "Epoch 269/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.8906\n",
            "Epoch 270/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.8213\n",
            "Epoch 271/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.8086\n",
            "Epoch 272/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.7725\n",
            "Epoch 273/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.7257\n",
            "Epoch 274/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.6938\n",
            "Epoch 275/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.6316\n",
            "Epoch 276/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.6261\n",
            "Epoch 277/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.6057\n",
            "Epoch 278/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.6061\n",
            "Epoch 279/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.5803\n",
            "Epoch 280/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 17.4638\n",
            "Epoch 281/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 17.5342\n",
            "Epoch 282/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.3722\n",
            "Epoch 283/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.4976\n",
            "Epoch 284/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.5042\n",
            "Epoch 285/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 17.3206\n",
            "Epoch 286/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.2902\n",
            "Epoch 287/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.2504\n",
            "Epoch 288/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.1601\n",
            "Epoch 289/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.1912\n",
            "Epoch 290/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.1088\n",
            "Epoch 291/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.0808\n",
            "Epoch 292/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.0400\n",
            "Epoch 293/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.0513\n",
            "Epoch 294/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.0968\n",
            "Epoch 295/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.0404\n",
            "Epoch 296/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 17.0259\n",
            "Epoch 297/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 16.8864\n",
            "Epoch 298/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 16.9340\n",
            "Epoch 299/300\n",
            "7/7 [==============================] - 0s 2ms/step - loss: 16.9191\n",
            "Epoch 300/300\n",
            "7/7 [==============================] - 0s 3ms/step - loss: 16.8254\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XvlzRAm7heKc",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 483
        },
        "outputId": "e5167be5-6c7c-49de-df1f-29ff4a8337fb"
      },
      "source": [
        "plt.plot(history.history['loss'])\n",
        "plt.show()"
      ],
      "execution_count": 126,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1152x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hA8Y46s6lLSf",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df[\"prediction\"] = model.predict(df["X"])"
      ],
      "execution_count": 127,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C6nXUw7mnO6E",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 483
        },
        "outputId": "0578e92b-0ab5-42bc-9b65-1cd2e7363584"
      },
      "source": [
        "plt.scatter(df[\"X\"], df[\"y\"])\n",
        "plt.plot(df[\"X\"], df[\"prediction\"], color='r')\n",
        "plt.show()"
      ],
      "execution_count": 130,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1152x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kBTw08jfq-eM",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 128,
      "outputs": []
    }
  ]
}