Recurrent_Neural_Network.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Prereq libraries\n",
    "import numpy as np\n",
    "import pandas as pb\n",
    "\n",
    "# Possibly useful TensorFlow and Keras\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Dropout, LSTM, CuDNNLSTM\n",
    "import tensorflow as tf\n",
    "\n",
    "# Dataset parameters\n",
    "TRAINING_SPLIT = 0.2\n",
    "FILE = 'min75_max5k_as_ids.csv'\n",
    "\n",
    "# Read in dataset from csv files\n",
    "dataset = pb.read_csv(FILE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Max Y Index: 130\n",
      "Max X Index: 3276\n",
      "X_train's Shape: 14270\n"
     ]
    }
   ],
   "source": [
    "from functs import npFloatArray, load_data\n",
    "\n",
    "# load features and labels based on those present in the CSV files\n",
    "features = np.array(dataset.copy())\n",
    "\n",
    "# Load Data\n",
    "X_train, X_test, y_train, y_test = load_data(dataset, \n",
    "                                             TRAINING_SPLIT)\n",
    "\n",
    "# Double check X_train's shape\n",
    "print(\"X_train's Shape: \" + str(X_train.shape[0]))\n",
    "\n",
    "# Set up dataset for model fitting\n",
    "num_rows, num_cols = X_train.shape\n",
    "X_train = np.reshape(X_train, (X_train.shape[0],\n",
    "                               1,\n",
    "                               X_train.shape[1]))\n",
    "X_test = np.reshape(X_test, (X_test.shape[0],\n",
    "                             1,\n",
    "                             X_test.shape[1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:Layer lstm will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
      "WARNING:tensorflow:Layer lstm_1 will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
      "Epoch 1/200\n",
      "446/446 [==============================] - 14s 26ms/step - loss: 0.0301 - accuracy: 0.2167 - precision: 0.7822 - root_mean_squared_error: 0.1735 - val_loss: 0.0291 - val_accuracy: 0.2710 - val_precision: 0.8062 - val_root_mean_squared_error: 0.1707\n",
      "Epoch 2/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0276 - accuracy: 0.3007 - precision: 0.8635 - root_mean_squared_error: 0.1663 - val_loss: 0.0287 - val_accuracy: 0.2478 - val_precision: 0.8184 - val_root_mean_squared_error: 0.1695\n",
      "Epoch 3/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0265 - accuracy: 0.3493 - precision: 0.9229 - root_mean_squared_error: 0.1628 - val_loss: 0.0286 - val_accuracy: 0.2803 - val_precision: 0.8055 - val_root_mean_squared_error: 0.1690\n",
      "Epoch 4/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0255 - accuracy: 0.3856 - precision: 0.9632 - root_mean_squared_error: 0.1598 - val_loss: 0.0286 - val_accuracy: 0.2654 - val_precision: 0.7835 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 5/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0249 - accuracy: 0.4087 - precision: 0.9796 - root_mean_squared_error: 0.1579 - val_loss: 0.0285 - val_accuracy: 0.2724 - val_precision: 0.8035 - val_root_mean_squared_error: 0.1688\n",
      "Epoch 6/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0245 - accuracy: 0.4255 - precision: 0.9889 - root_mean_squared_error: 0.1565 - val_loss: 0.0284 - val_accuracy: 0.2584 - val_precision: 0.7951 - val_root_mean_squared_error: 0.1684\n",
      "Epoch 7/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0242 - accuracy: 0.4324 - precision: 0.9911 - root_mean_squared_error: 0.1554 - val_loss: 0.0284 - val_accuracy: 0.2755 - val_precision: 0.8122 - val_root_mean_squared_error: 0.1684\n",
      "Epoch 8/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0239 - accuracy: 0.4371 - precision: 0.9931 - root_mean_squared_error: 0.1545 - val_loss: 0.0283 - val_accuracy: 0.2752 - val_precision: 0.8085 - val_root_mean_squared_error: 0.1682\n",
      "Epoch 9/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0236 - accuracy: 0.4403 - precision: 0.9955 - root_mean_squared_error: 0.1538 - val_loss: 0.0283 - val_accuracy: 0.2862 - val_precision: 0.8157 - val_root_mean_squared_error: 0.1681\n",
      "Epoch 10/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0235 - accuracy: 0.4452 - precision: 0.9970 - root_mean_squared_error: 0.1531 - val_loss: 0.0282 - val_accuracy: 0.2506 - val_precision: 0.8101 - val_root_mean_squared_error: 0.1679\n",
      "Epoch 11/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0233 - accuracy: 0.4461 - precision: 0.9954 - root_mean_squared_error: 0.1526 - val_loss: 0.0282 - val_accuracy: 0.2805 - val_precision: 0.7965 - val_root_mean_squared_error: 0.1680\n",
      "Epoch 12/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0232 - accuracy: 0.4463 - precision: 0.9961 - root_mean_squared_error: 0.1522 - val_loss: 0.0282 - val_accuracy: 0.2744 - val_precision: 0.8113 - val_root_mean_squared_error: 0.1680\n",
      "Epoch 13/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0231 - accuracy: 0.4481 - precision: 0.9960 - root_mean_squared_error: 0.1518 - val_loss: 0.0282 - val_accuracy: 0.2677 - val_precision: 0.7843 - val_root_mean_squared_error: 0.1679\n",
      "Epoch 14/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0230 - accuracy: 0.4462 - precision: 0.9964 - root_mean_squared_error: 0.1515 - val_loss: 0.0282 - val_accuracy: 0.2550 - val_precision: 0.8146 - val_root_mean_squared_error: 0.1681\n",
      "Epoch 15/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0229 - accuracy: 0.4481 - precision: 0.9969 - root_mean_squared_error: 0.1513 - val_loss: 0.0282 - val_accuracy: 0.2755 - val_precision: 0.7975 - val_root_mean_squared_error: 0.1680\n",
      "Epoch 16/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0228 - accuracy: 0.4502 - precision: 0.9965 - root_mean_squared_error: 0.1510 - val_loss: 0.0282 - val_accuracy: 0.2702 - val_precision: 0.8053 - val_root_mean_squared_error: 0.1678\n",
      "Epoch 17/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0228 - accuracy: 0.4468 - precision: 0.9967 - root_mean_squared_error: 0.1509 - val_loss: 0.0282 - val_accuracy: 0.2506 - val_precision: 0.7948 - val_root_mean_squared_error: 0.1679\n",
      "Epoch 18/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0227 - accuracy: 0.4485 - precision: 0.9966 - root_mean_squared_error: 0.1507 - val_loss: 0.0283 - val_accuracy: 0.2604 - val_precision: 0.7981 - val_root_mean_squared_error: 0.1682\n",
      "Epoch 19/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0227 - accuracy: 0.4521 - precision: 0.9969 - root_mean_squared_error: 0.1506 - val_loss: 0.0282 - val_accuracy: 0.2466 - val_precision: 0.8036 - val_root_mean_squared_error: 0.1681\n",
      "Epoch 20/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0226 - accuracy: 0.4498 - precision: 0.9965 - root_mean_squared_error: 0.1504 - val_loss: 0.0283 - val_accuracy: 0.2615 - val_precision: 0.8038 - val_root_mean_squared_error: 0.1681\n",
      "Epoch 21/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0226 - accuracy: 0.4495 - precision: 0.9964 - root_mean_squared_error: 0.1503 - val_loss: 0.0282 - val_accuracy: 0.2724 - val_precision: 0.7959 - val_root_mean_squared_error: 0.1680\n",
      "Epoch 22/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0226 - accuracy: 0.4468 - precision: 0.9973 - root_mean_squared_error: 0.1503 - val_loss: 0.0283 - val_accuracy: 0.2710 - val_precision: 0.8200 - val_root_mean_squared_error: 0.1681\n",
      "Epoch 23/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0226 - accuracy: 0.4436 - precision: 0.9964 - root_mean_squared_error: 0.1502 - val_loss: 0.0283 - val_accuracy: 0.2724 - val_precision: 0.7987 - val_root_mean_squared_error: 0.1682\n",
      "Epoch 24/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0225 - accuracy: 0.4425 - precision: 0.9963 - root_mean_squared_error: 0.1501 - val_loss: 0.0282 - val_accuracy: 0.2559 - val_precision: 0.7908 - val_root_mean_squared_error: 0.1681\n",
      "Epoch 25/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0225 - accuracy: 0.4434 - precision: 0.9974 - root_mean_squared_error: 0.1500 - val_loss: 0.0283 - val_accuracy: 0.2433 - val_precision: 0.7862 - val_root_mean_squared_error: 0.1681\n",
      "Epoch 26/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0225 - accuracy: 0.4422 - precision: 0.9973 - root_mean_squared_error: 0.1500 - val_loss: 0.0283 - val_accuracy: 0.2438 - val_precision: 0.7981 - val_root_mean_squared_error: 0.1682\n",
      "Epoch 27/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0225 - accuracy: 0.4486 - precision: 0.9971 - root_mean_squared_error: 0.1499 - val_loss: 0.0283 - val_accuracy: 0.2550 - val_precision: 0.8149 - val_root_mean_squared_error: 0.1683\n",
      "Epoch 28/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0225 - accuracy: 0.4423 - precision: 0.9965 - root_mean_squared_error: 0.1498 - val_loss: 0.0283 - val_accuracy: 0.2578 - val_precision: 0.8179 - val_root_mean_squared_error: 0.1682\n",
      "Epoch 29/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0224 - accuracy: 0.4499 - precision: 0.9968 - root_mean_squared_error: 0.1498 - val_loss: 0.0282 - val_accuracy: 0.2805 - val_precision: 0.8149 - val_root_mean_squared_error: 0.1680\n",
      "Epoch 30/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0224 - accuracy: 0.4334 - precision: 0.9985 - root_mean_squared_error: 0.1497 - val_loss: 0.0283 - val_accuracy: 0.2702 - val_precision: 0.8160 - val_root_mean_squared_error: 0.1683\n",
      "Epoch 31/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0224 - accuracy: 0.4406 - precision: 0.9969 - root_mean_squared_error: 0.1496 - val_loss: 0.0283 - val_accuracy: 0.2794 - val_precision: 0.7997 - val_root_mean_squared_error: 0.1683\n",
      "Epoch 32/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0224 - accuracy: 0.4419 - precision: 0.9981 - root_mean_squared_error: 0.1496 - val_loss: 0.0284 - val_accuracy: 0.2598 - val_precision: 0.8083 - val_root_mean_squared_error: 0.1684\n",
      "Epoch 33/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0224 - accuracy: 0.4463 - precision: 0.9965 - root_mean_squared_error: 0.1496 - val_loss: 0.0283 - val_accuracy: 0.2393 - val_precision: 0.7852 - val_root_mean_squared_error: 0.1683\n",
      "Epoch 34/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0224 - accuracy: 0.4393 - precision: 0.9968 - root_mean_squared_error: 0.1495 - val_loss: 0.0283 - val_accuracy: 0.2769 - val_precision: 0.7970 - val_root_mean_squared_error: 0.1682\n",
      "Epoch 35/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0223 - accuracy: 0.4399 - precision: 0.9961 - root_mean_squared_error: 0.1495 - val_loss: 0.0284 - val_accuracy: 0.2576 - val_precision: 0.8071 - val_root_mean_squared_error: 0.1684\n",
      "Epoch 36/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0223 - accuracy: 0.4355 - precision: 0.9969 - root_mean_squared_error: 0.1495 - val_loss: 0.0283 - val_accuracy: 0.2539 - val_precision: 0.7976 - val_root_mean_squared_error: 0.1682\n",
      "Epoch 37/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0223 - accuracy: 0.4357 - precision: 0.9969 - root_mean_squared_error: 0.1494 - val_loss: 0.0283 - val_accuracy: 0.2416 - val_precision: 0.8081 - val_root_mean_squared_error: 0.1683\n",
      "Epoch 38/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0223 - accuracy: 0.4319 - precision: 0.9967 - root_mean_squared_error: 0.1494 - val_loss: 0.0283 - val_accuracy: 0.2494 - val_precision: 0.7830 - val_root_mean_squared_error: 0.1683\n",
      "Epoch 39/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0223 - accuracy: 0.4397 - precision: 0.9969 - root_mean_squared_error: 0.1493 - val_loss: 0.0283 - val_accuracy: 0.2674 - val_precision: 0.7959 - val_root_mean_squared_error: 0.1682\n",
      "Epoch 40/200\n",
      "446/446 [==============================] - 12s 26ms/step - loss: 0.0223 - accuracy: 0.4395 - precision: 0.9974 - root_mean_squared_error: 0.1493 - val_loss: 0.0283 - val_accuracy: 0.2497 - val_precision: 0.7911 - val_root_mean_squared_error: 0.1683\n",
      "Epoch 41/200\n",
      "446/446 [==============================] - 12s 26ms/step - loss: 0.0223 - accuracy: 0.4367 - precision: 0.9970 - root_mean_squared_error: 0.1492 - val_loss: 0.0283 - val_accuracy: 0.2318 - val_precision: 0.7810 - val_root_mean_squared_error: 0.1683\n",
      "Epoch 42/200\n",
      "446/446 [==============================] - 12s 26ms/step - loss: 0.0223 - accuracy: 0.4359 - precision: 0.9971 - root_mean_squared_error: 0.1492 - val_loss: 0.0284 - val_accuracy: 0.2550 - val_precision: 0.7964 - val_root_mean_squared_error: 0.1684\n",
      "Epoch 43/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0222 - accuracy: 0.4348 - precision: 0.9969 - root_mean_squared_error: 0.1491 - val_loss: 0.0282 - val_accuracy: 0.2346 - val_precision: 0.7705 - val_root_mean_squared_error: 0.1681\n",
      "Epoch 44/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0222 - accuracy: 0.4336 - precision: 0.9966 - root_mean_squared_error: 0.1491 - val_loss: 0.0283 - val_accuracy: 0.2483 - val_precision: 0.8103 - val_root_mean_squared_error: 0.1682\n",
      "Epoch 45/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0222 - accuracy: 0.4256 - precision: 0.9968 - root_mean_squared_error: 0.1490 - val_loss: 0.0284 - val_accuracy: 0.2419 - val_precision: 0.7925 - val_root_mean_squared_error: 0.1684\n",
      "Epoch 46/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0222 - accuracy: 0.4319 - precision: 0.9965 - root_mean_squared_error: 0.1490 - val_loss: 0.0283 - val_accuracy: 0.2329 - val_precision: 0.7874 - val_root_mean_squared_error: 0.1683\n",
      "Epoch 47/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0222 - accuracy: 0.4294 - precision: 0.9966 - root_mean_squared_error: 0.1490 - val_loss: 0.0283 - val_accuracy: 0.2466 - val_precision: 0.7850 - val_root_mean_squared_error: 0.1684\n",
      "Epoch 48/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0222 - accuracy: 0.4371 - precision: 0.9962 - root_mean_squared_error: 0.1489 - val_loss: 0.0284 - val_accuracy: 0.2326 - val_precision: 0.7951 - val_root_mean_squared_error: 0.1686\n",
      "Epoch 49/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0222 - accuracy: 0.4266 - precision: 0.9967 - root_mean_squared_error: 0.1489 - val_loss: 0.0284 - val_accuracy: 0.2351 - val_precision: 0.7977 - val_root_mean_squared_error: 0.1685\n",
      "Epoch 50/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0222 - accuracy: 0.4237 - precision: 0.9964 - root_mean_squared_error: 0.1489 - val_loss: 0.0284 - val_accuracy: 0.2480 - val_precision: 0.7783 - val_root_mean_squared_error: 0.1684\n",
      "Epoch 51/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0222 - accuracy: 0.4302 - precision: 0.9966 - root_mean_squared_error: 0.1489 - val_loss: 0.0283 - val_accuracy: 0.2402 - val_precision: 0.8161 - val_root_mean_squared_error: 0.1683\n",
      "Epoch 52/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0222 - accuracy: 0.4224 - precision: 0.9969 - root_mean_squared_error: 0.1488 - val_loss: 0.0284 - val_accuracy: 0.2349 - val_precision: 0.7963 - val_root_mean_squared_error: 0.1686\n",
      "Epoch 53/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0221 - accuracy: 0.4273 - precision: 0.9970 - root_mean_squared_error: 0.1488 - val_loss: 0.0283 - val_accuracy: 0.2161 - val_precision: 0.7846 - val_root_mean_squared_error: 0.1683\n",
      "Epoch 54/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0221 - accuracy: 0.4159 - precision: 0.9964 - root_mean_squared_error: 0.1487 - val_loss: 0.0284 - val_accuracy: 0.2363 - val_precision: 0.7919 - val_root_mean_squared_error: 0.1684\n",
      "Epoch 55/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0221 - accuracy: 0.4233 - precision: 0.9969 - root_mean_squared_error: 0.1487 - val_loss: 0.0283 - val_accuracy: 0.2424 - val_precision: 0.8113 - val_root_mean_squared_error: 0.1683\n",
      "Epoch 56/200\n",
      "446/446 [==============================] - 12s 26ms/step - loss: 0.0221 - accuracy: 0.4221 - precision: 0.9968 - root_mean_squared_error: 0.1486 - val_loss: 0.0284 - val_accuracy: 0.2290 - val_precision: 0.7997 - val_root_mean_squared_error: 0.1684\n",
      "Epoch 57/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0221 - accuracy: 0.4215 - precision: 0.9968 - root_mean_squared_error: 0.1486 - val_loss: 0.0284 - val_accuracy: 0.2503 - val_precision: 0.7914 - val_root_mean_squared_error: 0.1685\n",
      "Epoch 58/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0221 - accuracy: 0.4198 - precision: 0.9968 - root_mean_squared_error: 0.1486 - val_loss: 0.0284 - val_accuracy: 0.2256 - val_precision: 0.7982 - val_root_mean_squared_error: 0.1684\n",
      "Epoch 59/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0221 - accuracy: 0.4186 - precision: 0.9965 - root_mean_squared_error: 0.1486 - val_loss: 0.0284 - val_accuracy: 0.2231 - val_precision: 0.7903 - val_root_mean_squared_error: 0.1684\n",
      "Epoch 60/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0221 - accuracy: 0.4233 - precision: 0.9964 - root_mean_squared_error: 0.1485 - val_loss: 0.0284 - val_accuracy: 0.2469 - val_precision: 0.8061 - val_root_mean_squared_error: 0.1685\n",
      "Epoch 61/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0221 - accuracy: 0.4174 - precision: 0.9968 - root_mean_squared_error: 0.1485 - val_loss: 0.0283 - val_accuracy: 0.2270 - val_precision: 0.7807 - val_root_mean_squared_error: 0.1683\n",
      "Epoch 62/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0221 - accuracy: 0.4149 - precision: 0.9965 - root_mean_squared_error: 0.1485 - val_loss: 0.0283 - val_accuracy: 0.2612 - val_precision: 0.7838 - val_root_mean_squared_error: 0.1682\n",
      "Epoch 63/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0220 - accuracy: 0.4206 - precision: 0.9966 - root_mean_squared_error: 0.1485 - val_loss: 0.0283 - val_accuracy: 0.2413 - val_precision: 0.7895 - val_root_mean_squared_error: 0.1683\n",
      "Epoch 64/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0220 - accuracy: 0.4177 - precision: 0.9968 - root_mean_squared_error: 0.1485 - val_loss: 0.0283 - val_accuracy: 0.2337 - val_precision: 0.7967 - val_root_mean_squared_error: 0.1683\n",
      "Epoch 65/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0220 - accuracy: 0.4163 - precision: 0.9963 - root_mean_squared_error: 0.1485 - val_loss: 0.0284 - val_accuracy: 0.2396 - val_precision: 0.7897 - val_root_mean_squared_error: 0.1685\n",
      "Epoch 66/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0220 - accuracy: 0.4180 - precision: 0.9967 - root_mean_squared_error: 0.1484 - val_loss: 0.0284 - val_accuracy: 0.2326 - val_precision: 0.7864 - val_root_mean_squared_error: 0.1684\n",
      "Epoch 67/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0220 - accuracy: 0.4211 - precision: 0.9965 - root_mean_squared_error: 0.1484 - val_loss: 0.0284 - val_accuracy: 0.2539 - val_precision: 0.8118 - val_root_mean_squared_error: 0.1684\n",
      "Epoch 68/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0220 - accuracy: 0.4132 - precision: 0.9966 - root_mean_squared_error: 0.1484 - val_loss: 0.0284 - val_accuracy: 0.2290 - val_precision: 0.7748 - val_root_mean_squared_error: 0.1685\n",
      "Epoch 69/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0220 - accuracy: 0.4185 - precision: 0.9965 - root_mean_squared_error: 0.1484 - val_loss: 0.0284 - val_accuracy: 0.2276 - val_precision: 0.7769 - val_root_mean_squared_error: 0.1685\n",
      "Epoch 70/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0220 - accuracy: 0.4184 - precision: 0.9965 - root_mean_squared_error: 0.1484 - val_loss: 0.0284 - val_accuracy: 0.2469 - val_precision: 0.7878 - val_root_mean_squared_error: 0.1685\n",
      "Epoch 71/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0220 - accuracy: 0.4221 - precision: 0.9974 - root_mean_squared_error: 0.1484 - val_loss: 0.0284 - val_accuracy: 0.2402 - val_precision: 0.7851 - val_root_mean_squared_error: 0.1685\n",
      "Epoch 72/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0220 - accuracy: 0.4178 - precision: 0.9966 - root_mean_squared_error: 0.1483 - val_loss: 0.0284 - val_accuracy: 0.2379 - val_precision: 0.7932 - val_root_mean_squared_error: 0.1684\n",
      "Epoch 73/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0220 - accuracy: 0.4179 - precision: 0.9967 - root_mean_squared_error: 0.1483 - val_loss: 0.0284 - val_accuracy: 0.2550 - val_precision: 0.7968 - val_root_mean_squared_error: 0.1685\n",
      "Epoch 74/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0220 - accuracy: 0.4134 - precision: 0.9967 - root_mean_squared_error: 0.1483 - val_loss: 0.0285 - val_accuracy: 0.2346 - val_precision: 0.7925 - val_root_mean_squared_error: 0.1687\n",
      "Epoch 75/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0220 - accuracy: 0.4135 - precision: 0.9967 - root_mean_squared_error: 0.1483 - val_loss: 0.0284 - val_accuracy: 0.2382 - val_precision: 0.7782 - val_root_mean_squared_error: 0.1685\n",
      "Epoch 76/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0220 - accuracy: 0.4153 - precision: 0.9966 - root_mean_squared_error: 0.1483 - val_loss: 0.0284 - val_accuracy: 0.2166 - val_precision: 0.7721 - val_root_mean_squared_error: 0.1684\n",
      "Epoch 77/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0220 - accuracy: 0.4192 - precision: 0.9971 - root_mean_squared_error: 0.1483 - val_loss: 0.0284 - val_accuracy: 0.2290 - val_precision: 0.7759 - val_root_mean_squared_error: 0.1687\n",
      "Epoch 78/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0220 - accuracy: 0.4128 - precision: 0.9967 - root_mean_squared_error: 0.1483 - val_loss: 0.0285 - val_accuracy: 0.2192 - val_precision: 0.7865 - val_root_mean_squared_error: 0.1687\n",
      "Epoch 79/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0220 - accuracy: 0.4186 - precision: 0.9966 - root_mean_squared_error: 0.1482 - val_loss: 0.0284 - val_accuracy: 0.2405 - val_precision: 0.7923 - val_root_mean_squared_error: 0.1684\n",
      "Epoch 80/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0220 - accuracy: 0.4095 - precision: 0.9967 - root_mean_squared_error: 0.1482 - val_loss: 0.0284 - val_accuracy: 0.2385 - val_precision: 0.7769 - val_root_mean_squared_error: 0.1685\n",
      "Epoch 81/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0220 - accuracy: 0.4167 - precision: 0.9964 - root_mean_squared_error: 0.1482 - val_loss: 0.0285 - val_accuracy: 0.2315 - val_precision: 0.7773 - val_root_mean_squared_error: 0.1687\n",
      "Epoch 82/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0220 - accuracy: 0.4170 - precision: 0.9963 - root_mean_squared_error: 0.1482 - val_loss: 0.0284 - val_accuracy: 0.2422 - val_precision: 0.7830 - val_root_mean_squared_error: 0.1686\n",
      "Epoch 83/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0220 - accuracy: 0.4102 - precision: 0.9964 - root_mean_squared_error: 0.1482 - val_loss: 0.0284 - val_accuracy: 0.2696 - val_precision: 0.7788 - val_root_mean_squared_error: 0.1685\n",
      "Epoch 84/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0220 - accuracy: 0.4196 - precision: 0.9968 - root_mean_squared_error: 0.1482 - val_loss: 0.0284 - val_accuracy: 0.2444 - val_precision: 0.7917 - val_root_mean_squared_error: 0.1686\n",
      "Epoch 85/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0220 - accuracy: 0.4114 - precision: 0.9967 - root_mean_squared_error: 0.1482 - val_loss: 0.0284 - val_accuracy: 0.2379 - val_precision: 0.7917 - val_root_mean_squared_error: 0.1686\n",
      "Epoch 86/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0220 - accuracy: 0.4151 - precision: 0.9969 - root_mean_squared_error: 0.1482 - val_loss: 0.0285 - val_accuracy: 0.2231 - val_precision: 0.7917 - val_root_mean_squared_error: 0.1687\n",
      "Epoch 87/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0219 - accuracy: 0.4179 - precision: 0.9967 - root_mean_squared_error: 0.1482 - val_loss: 0.0285 - val_accuracy: 0.2578 - val_precision: 0.7982 - val_root_mean_squared_error: 0.1687\n",
      "Epoch 88/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4158 - precision: 0.9967 - root_mean_squared_error: 0.1481 - val_loss: 0.0285 - val_accuracy: 0.2567 - val_precision: 0.7933 - val_root_mean_squared_error: 0.1688\n",
      "Epoch 89/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4175 - precision: 0.9961 - root_mean_squared_error: 0.1481 - val_loss: 0.0285 - val_accuracy: 0.2545 - val_precision: 0.8030 - val_root_mean_squared_error: 0.1688\n",
      "Epoch 90/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4146 - precision: 0.9972 - root_mean_squared_error: 0.1481 - val_loss: 0.0285 - val_accuracy: 0.2497 - val_precision: 0.7892 - val_root_mean_squared_error: 0.1688\n",
      "Epoch 91/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4142 - precision: 0.9965 - root_mean_squared_error: 0.1481 - val_loss: 0.0285 - val_accuracy: 0.2068 - val_precision: 0.7910 - val_root_mean_squared_error: 0.1689\n",
      "Epoch 92/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4165 - precision: 0.9964 - root_mean_squared_error: 0.1481 - val_loss: 0.0285 - val_accuracy: 0.2351 - val_precision: 0.7917 - val_root_mean_squared_error: 0.1689\n",
      "Epoch 93/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0219 - accuracy: 0.4114 - precision: 0.9965 - root_mean_squared_error: 0.1481 - val_loss: 0.0285 - val_accuracy: 0.2461 - val_precision: 0.8036 - val_root_mean_squared_error: 0.1687\n",
      "Epoch 94/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4082 - precision: 0.9962 - root_mean_squared_error: 0.1481 - val_loss: 0.0285 - val_accuracy: 0.2379 - val_precision: 0.8012 - val_root_mean_squared_error: 0.1688\n",
      "Epoch 95/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0219 - accuracy: 0.4144 - precision: 0.9966 - root_mean_squared_error: 0.1481 - val_loss: 0.0285 - val_accuracy: 0.2265 - val_precision: 0.7982 - val_root_mean_squared_error: 0.1687\n",
      "Epoch 96/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4097 - precision: 0.9966 - root_mean_squared_error: 0.1481 - val_loss: 0.0284 - val_accuracy: 0.2433 - val_precision: 0.7988 - val_root_mean_squared_error: 0.1687\n",
      "Epoch 97/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0219 - accuracy: 0.4112 - precision: 0.9965 - root_mean_squared_error: 0.1480 - val_loss: 0.0285 - val_accuracy: 0.2514 - val_precision: 0.8052 - val_root_mean_squared_error: 0.1688\n",
      "Epoch 98/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4133 - precision: 0.9965 - root_mean_squared_error: 0.1480 - val_loss: 0.0285 - val_accuracy: 0.2234 - val_precision: 0.7974 - val_root_mean_squared_error: 0.1689\n",
      "Epoch 99/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4121 - precision: 0.9966 - root_mean_squared_error: 0.1480 - val_loss: 0.0284 - val_accuracy: 0.2595 - val_precision: 0.7903 - val_root_mean_squared_error: 0.1686\n",
      "Epoch 100/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4142 - precision: 0.9965 - root_mean_squared_error: 0.1480 - val_loss: 0.0285 - val_accuracy: 0.2354 - val_precision: 0.8028 - val_root_mean_squared_error: 0.1688\n",
      "Epoch 101/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4127 - precision: 0.9965 - root_mean_squared_error: 0.1480 - val_loss: 0.0285 - val_accuracy: 0.2307 - val_precision: 0.7821 - val_root_mean_squared_error: 0.1689\n",
      "Epoch 102/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4135 - precision: 0.9966 - root_mean_squared_error: 0.1480 - val_loss: 0.0284 - val_accuracy: 0.2270 - val_precision: 0.7955 - val_root_mean_squared_error: 0.1686\n",
      "Epoch 103/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4134 - precision: 0.9966 - root_mean_squared_error: 0.1480 - val_loss: 0.0284 - val_accuracy: 0.2618 - val_precision: 0.7982 - val_root_mean_squared_error: 0.1685\n",
      "Epoch 104/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4153 - precision: 0.9966 - root_mean_squared_error: 0.1480 - val_loss: 0.0284 - val_accuracy: 0.2287 - val_precision: 0.8077 - val_root_mean_squared_error: 0.1686\n",
      "Epoch 105/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4093 - precision: 0.9966 - root_mean_squared_error: 0.1480 - val_loss: 0.0284 - val_accuracy: 0.2195 - val_precision: 0.7982 - val_root_mean_squared_error: 0.1686\n",
      "Epoch 106/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0219 - accuracy: 0.4081 - precision: 0.9969 - root_mean_squared_error: 0.1480 - val_loss: 0.0285 - val_accuracy: 0.2242 - val_precision: 0.8132 - val_root_mean_squared_error: 0.1688\n",
      "Epoch 107/200\n",
      "446/446 [==============================] - 12s 26ms/step - loss: 0.0219 - accuracy: 0.4055 - precision: 0.9966 - root_mean_squared_error: 0.1480 - val_loss: 0.0285 - val_accuracy: 0.2239 - val_precision: 0.8020 - val_root_mean_squared_error: 0.1688\n",
      "Epoch 108/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0219 - accuracy: 0.4148 - precision: 0.9969 - root_mean_squared_error: 0.1480 - val_loss: 0.0285 - val_accuracy: 0.2424 - val_precision: 0.7797 - val_root_mean_squared_error: 0.1689\n",
      "Epoch 109/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0219 - accuracy: 0.4088 - precision: 0.9966 - root_mean_squared_error: 0.1480 - val_loss: 0.0285 - val_accuracy: 0.2399 - val_precision: 0.8089 - val_root_mean_squared_error: 0.1687\n",
      "Epoch 110/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4150 - precision: 0.9966 - root_mean_squared_error: 0.1479 - val_loss: 0.0285 - val_accuracy: 0.2430 - val_precision: 0.7991 - val_root_mean_squared_error: 0.1687\n",
      "Epoch 111/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4104 - precision: 0.9966 - root_mean_squared_error: 0.1479 - val_loss: 0.0285 - val_accuracy: 0.2469 - val_precision: 0.7824 - val_root_mean_squared_error: 0.1690\n",
      "Epoch 112/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4170 - precision: 0.9964 - root_mean_squared_error: 0.1479 - val_loss: 0.0285 - val_accuracy: 0.2307 - val_precision: 0.7955 - val_root_mean_squared_error: 0.1688\n",
      "Epoch 113/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4163 - precision: 0.9964 - root_mean_squared_error: 0.1479 - val_loss: 0.0285 - val_accuracy: 0.2553 - val_precision: 0.7863 - val_root_mean_squared_error: 0.1687\n",
      "Epoch 114/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0219 - accuracy: 0.4109 - precision: 0.9966 - root_mean_squared_error: 0.1479 - val_loss: 0.0285 - val_accuracy: 0.2410 - val_precision: 0.8042 - val_root_mean_squared_error: 0.1689\n",
      "Epoch 115/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4131 - precision: 0.9966 - root_mean_squared_error: 0.1479 - val_loss: 0.0285 - val_accuracy: 0.2276 - val_precision: 0.7977 - val_root_mean_squared_error: 0.1688\n",
      "Epoch 116/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4160 - precision: 0.9964 - root_mean_squared_error: 0.1479 - val_loss: 0.0285 - val_accuracy: 0.2357 - val_precision: 0.7927 - val_root_mean_squared_error: 0.1688\n",
      "Epoch 117/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4125 - precision: 0.9966 - root_mean_squared_error: 0.1479 - val_loss: 0.0286 - val_accuracy: 0.2251 - val_precision: 0.7925 - val_root_mean_squared_error: 0.1690\n",
      "Epoch 118/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4179 - precision: 0.9966 - root_mean_squared_error: 0.1479 - val_loss: 0.0285 - val_accuracy: 0.2298 - val_precision: 0.7947 - val_root_mean_squared_error: 0.1689\n",
      "Epoch 119/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4071 - precision: 0.9964 - root_mean_squared_error: 0.1479 - val_loss: 0.0286 - val_accuracy: 0.2343 - val_precision: 0.7841 - val_root_mean_squared_error: 0.1690\n",
      "Epoch 120/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0219 - accuracy: 0.4141 - precision: 0.9964 - root_mean_squared_error: 0.1479 - val_loss: 0.0285 - val_accuracy: 0.2427 - val_precision: 0.7952 - val_root_mean_squared_error: 0.1688\n",
      "Epoch 121/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0219 - accuracy: 0.4152 - precision: 0.9964 - root_mean_squared_error: 0.1479 - val_loss: 0.0284 - val_accuracy: 0.2548 - val_precision: 0.7994 - val_root_mean_squared_error: 0.1687\n",
      "Epoch 122/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0219 - accuracy: 0.4141 - precision: 0.9966 - root_mean_squared_error: 0.1479 - val_loss: 0.0285 - val_accuracy: 0.2349 - val_precision: 0.7860 - val_root_mean_squared_error: 0.1687\n",
      "Epoch 123/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4122 - precision: 0.9966 - root_mean_squared_error: 0.1479 - val_loss: 0.0285 - val_accuracy: 0.2466 - val_precision: 0.7827 - val_root_mean_squared_error: 0.1687\n",
      "Epoch 124/200\n",
      "446/446 [==============================] - 12s 26ms/step - loss: 0.0219 - accuracy: 0.4104 - precision: 0.9966 - root_mean_squared_error: 0.1479 - val_loss: 0.0285 - val_accuracy: 0.2251 - val_precision: 0.7846 - val_root_mean_squared_error: 0.1689\n",
      "Epoch 125/200\n",
      "446/446 [==============================] - 12s 26ms/step - loss: 0.0219 - accuracy: 0.4144 - precision: 0.9965 - root_mean_squared_error: 0.1479 - val_loss: 0.0285 - val_accuracy: 0.2430 - val_precision: 0.7882 - val_root_mean_squared_error: 0.1687\n",
      "Epoch 126/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4140 - precision: 0.9965 - root_mean_squared_error: 0.1479 - val_loss: 0.0285 - val_accuracy: 0.2239 - val_precision: 0.7834 - val_root_mean_squared_error: 0.1687\n",
      "Epoch 127/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4085 - precision: 0.9965 - root_mean_squared_error: 0.1479 - val_loss: 0.0285 - val_accuracy: 0.2419 - val_precision: 0.7915 - val_root_mean_squared_error: 0.1689\n",
      "Epoch 128/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4184 - precision: 0.9964 - root_mean_squared_error: 0.1479 - val_loss: 0.0286 - val_accuracy: 0.2267 - val_precision: 0.7913 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 129/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4167 - precision: 0.9963 - root_mean_squared_error: 0.1479 - val_loss: 0.0285 - val_accuracy: 0.2461 - val_precision: 0.7844 - val_root_mean_squared_error: 0.1690\n",
      "Epoch 130/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4128 - precision: 0.9967 - root_mean_squared_error: 0.1479 - val_loss: 0.0285 - val_accuracy: 0.2332 - val_precision: 0.7862 - val_root_mean_squared_error: 0.1689\n",
      "Epoch 131/200\n",
      "446/446 [==============================] - 12s 26ms/step - loss: 0.0219 - accuracy: 0.4132 - precision: 0.9965 - root_mean_squared_error: 0.1479 - val_loss: 0.0286 - val_accuracy: 0.2368 - val_precision: 0.7672 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 132/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4121 - precision: 0.9967 - root_mean_squared_error: 0.1478 - val_loss: 0.0286 - val_accuracy: 0.2396 - val_precision: 0.7718 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 133/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4104 - precision: 0.9964 - root_mean_squared_error: 0.1478 - val_loss: 0.0286 - val_accuracy: 0.2377 - val_precision: 0.7815 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 134/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0219 - accuracy: 0.4144 - precision: 0.9965 - root_mean_squared_error: 0.1478 - val_loss: 0.0285 - val_accuracy: 0.2517 - val_precision: 0.7834 - val_root_mean_squared_error: 0.1689\n",
      "Epoch 135/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4121 - precision: 0.9964 - root_mean_squared_error: 0.1478 - val_loss: 0.0286 - val_accuracy: 0.2186 - val_precision: 0.7795 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 136/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4127 - precision: 0.9964 - root_mean_squared_error: 0.1478 - val_loss: 0.0286 - val_accuracy: 0.2500 - val_precision: 0.7867 - val_root_mean_squared_error: 0.1690\n",
      "Epoch 137/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0219 - accuracy: 0.4153 - precision: 0.9965 - root_mean_squared_error: 0.1478 - val_loss: 0.0285 - val_accuracy: 0.2287 - val_precision: 0.7686 - val_root_mean_squared_error: 0.1689\n",
      "Epoch 138/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4091 - precision: 0.9964 - root_mean_squared_error: 0.1478 - val_loss: 0.0286 - val_accuracy: 0.2461 - val_precision: 0.7879 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 139/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0218 - accuracy: 0.4085 - precision: 0.9965 - root_mean_squared_error: 0.1478 - val_loss: 0.0286 - val_accuracy: 0.2391 - val_precision: 0.7825 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 140/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4064 - precision: 0.9962 - root_mean_squared_error: 0.1478 - val_loss: 0.0285 - val_accuracy: 0.2385 - val_precision: 0.7829 - val_root_mean_squared_error: 0.1690\n",
      "Epoch 141/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0218 - accuracy: 0.4105 - precision: 0.9964 - root_mean_squared_error: 0.1478 - val_loss: 0.0286 - val_accuracy: 0.2332 - val_precision: 0.7770 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 142/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4101 - precision: 0.9964 - root_mean_squared_error: 0.1478 - val_loss: 0.0285 - val_accuracy: 0.2450 - val_precision: 0.7769 - val_root_mean_squared_error: 0.1689\n",
      "Epoch 143/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0218 - accuracy: 0.4117 - precision: 0.9968 - root_mean_squared_error: 0.1478 - val_loss: 0.0286 - val_accuracy: 0.2702 - val_precision: 0.7836 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 144/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4134 - precision: 0.9965 - root_mean_squared_error: 0.1478 - val_loss: 0.0286 - val_accuracy: 0.2388 - val_precision: 0.7839 - val_root_mean_squared_error: 0.1690\n",
      "Epoch 145/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4105 - precision: 0.9965 - root_mean_squared_error: 0.1478 - val_loss: 0.0286 - val_accuracy: 0.2388 - val_precision: 0.7814 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 146/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4160 - precision: 0.9964 - root_mean_squared_error: 0.1478 - val_loss: 0.0286 - val_accuracy: 0.2444 - val_precision: 0.7874 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 147/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4123 - precision: 0.9965 - root_mean_squared_error: 0.1478 - val_loss: 0.0286 - val_accuracy: 0.2340 - val_precision: 0.7869 - val_root_mean_squared_error: 0.1692\n",
      "Epoch 148/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4102 - precision: 0.9965 - root_mean_squared_error: 0.1478 - val_loss: 0.0285 - val_accuracy: 0.2402 - val_precision: 0.7761 - val_root_mean_squared_error: 0.1688\n",
      "Epoch 149/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0218 - accuracy: 0.4153 - precision: 0.9964 - root_mean_squared_error: 0.1478 - val_loss: 0.0286 - val_accuracy: 0.2391 - val_precision: 0.7874 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 150/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4152 - precision: 0.9965 - root_mean_squared_error: 0.1478 - val_loss: 0.0286 - val_accuracy: 0.2556 - val_precision: 0.7861 - val_root_mean_squared_error: 0.1692\n",
      "Epoch 151/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0218 - accuracy: 0.4015 - precision: 0.9962 - root_mean_squared_error: 0.1478 - val_loss: 0.0286 - val_accuracy: 0.2559 - val_precision: 0.7816 - val_root_mean_squared_error: 0.1692\n",
      "Epoch 152/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0218 - accuracy: 0.4160 - precision: 0.9965 - root_mean_squared_error: 0.1478 - val_loss: 0.0286 - val_accuracy: 0.2357 - val_precision: 0.7773 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 153/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0218 - accuracy: 0.4142 - precision: 0.9965 - root_mean_squared_error: 0.1478 - val_loss: 0.0287 - val_accuracy: 0.2480 - val_precision: 0.7926 - val_root_mean_squared_error: 0.1693\n",
      "Epoch 154/200\n",
      "446/446 [==============================] - 12s 26ms/step - loss: 0.0218 - accuracy: 0.4100 - precision: 0.9962 - root_mean_squared_error: 0.1478 - val_loss: 0.0286 - val_accuracy: 0.2570 - val_precision: 0.7823 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 155/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0218 - accuracy: 0.4155 - precision: 0.9964 - root_mean_squared_error: 0.1478 - val_loss: 0.0285 - val_accuracy: 0.2419 - val_precision: 0.7740 - val_root_mean_squared_error: 0.1690\n",
      "Epoch 156/200\n",
      "446/446 [==============================] - 12s 26ms/step - loss: 0.0218 - accuracy: 0.4162 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2436 - val_precision: 0.7769 - val_root_mean_squared_error: 0.1690\n",
      "Epoch 157/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4128 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2436 - val_precision: 0.7917 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 158/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0218 - accuracy: 0.4145 - precision: 0.9964 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2346 - val_precision: 0.7852 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 159/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4117 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2391 - val_precision: 0.7856 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 160/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4071 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2374 - val_precision: 0.8095 - val_root_mean_squared_error: 0.1692\n",
      "Epoch 161/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4137 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2231 - val_precision: 0.8021 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 162/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4135 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0285 - val_accuracy: 0.2259 - val_precision: 0.7767 - val_root_mean_squared_error: 0.1689\n",
      "Epoch 163/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4132 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2444 - val_precision: 0.8048 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 164/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4120 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0285 - val_accuracy: 0.2332 - val_precision: 0.7926 - val_root_mean_squared_error: 0.1689\n",
      "Epoch 165/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4078 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2304 - val_precision: 0.8007 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 166/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4101 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2478 - val_precision: 0.7947 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 167/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4090 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2374 - val_precision: 0.8000 - val_root_mean_squared_error: 0.1690\n",
      "Epoch 168/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4091 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0285 - val_accuracy: 0.2396 - val_precision: 0.7874 - val_root_mean_squared_error: 0.1689\n",
      "Epoch 169/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4088 - precision: 0.9964 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2357 - val_precision: 0.7963 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 170/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4114 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2422 - val_precision: 0.7857 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 171/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4172 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2486 - val_precision: 0.7877 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 172/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4118 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2323 - val_precision: 0.7842 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 173/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4163 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2349 - val_precision: 0.7649 - val_root_mean_squared_error: 0.1690\n",
      "Epoch 174/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4180 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2528 - val_precision: 0.7829 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 175/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4108 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2295 - val_precision: 0.7702 - val_root_mean_squared_error: 0.1690\n",
      "Epoch 176/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4132 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2534 - val_precision: 0.7955 - val_root_mean_squared_error: 0.1692\n",
      "Epoch 177/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4104 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2363 - val_precision: 0.7903 - val_root_mean_squared_error: 0.1692\n",
      "Epoch 178/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4071 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2382 - val_precision: 0.7806 - val_root_mean_squared_error: 0.1692\n",
      "Epoch 179/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4160 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2545 - val_precision: 0.7857 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 180/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4142 - precision: 0.9963 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2539 - val_precision: 0.7848 - val_root_mean_squared_error: 0.1692\n",
      "Epoch 181/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4117 - precision: 0.9963 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2461 - val_precision: 0.7687 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 182/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4107 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0287 - val_accuracy: 0.2433 - val_precision: 0.7927 - val_root_mean_squared_error: 0.1693\n",
      "Epoch 183/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4116 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2377 - val_precision: 0.7993 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 184/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4113 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2391 - val_precision: 0.7749 - val_root_mean_squared_error: 0.1692\n",
      "Epoch 185/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4064 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2489 - val_precision: 0.7821 - val_root_mean_squared_error: 0.1690\n",
      "Epoch 186/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4157 - precision: 0.9962 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2584 - val_precision: 0.7849 - val_root_mean_squared_error: 0.1690\n",
      "Epoch 187/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4064 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2441 - val_precision: 0.7903 - val_root_mean_squared_error: 0.1692\n",
      "Epoch 188/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4095 - precision: 0.9966 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2295 - val_precision: 0.7858 - val_root_mean_squared_error: 0.1692\n",
      "Epoch 189/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4070 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2399 - val_precision: 0.7896 - val_root_mean_squared_error: 0.1690\n",
      "Epoch 190/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4152 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2438 - val_precision: 0.7897 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 191/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4138 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0285 - val_accuracy: 0.2393 - val_precision: 0.7835 - val_root_mean_squared_error: 0.1689\n",
      "Epoch 192/200\n",
      "446/446 [==============================] - 11s 26ms/step - loss: 0.0218 - accuracy: 0.4043 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0285 - val_accuracy: 0.2595 - val_precision: 0.7792 - val_root_mean_squared_error: 0.1690\n",
      "Epoch 193/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4149 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2396 - val_precision: 0.7807 - val_root_mean_squared_error: 0.1690\n",
      "Epoch 194/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4107 - precision: 0.9966 - root_mean_squared_error: 0.1477 - val_loss: 0.0285 - val_accuracy: 0.2335 - val_precision: 0.7832 - val_root_mean_squared_error: 0.1689\n",
      "Epoch 195/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4122 - precision: 0.9963 - root_mean_squared_error: 0.1477 - val_loss: 0.0285 - val_accuracy: 0.2584 - val_precision: 0.7908 - val_root_mean_squared_error: 0.1689\n",
      "Epoch 196/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4121 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0286 - val_accuracy: 0.2438 - val_precision: 0.7915 - val_root_mean_squared_error: 0.1690\n",
      "Epoch 197/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4150 - precision: 0.9965 - root_mean_squared_error: 0.1477 - val_loss: 0.0285 - val_accuracy: 0.2424 - val_precision: 0.7841 - val_root_mean_squared_error: 0.1689\n",
      "Epoch 198/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4145 - precision: 0.9965 - root_mean_squared_error: 0.1476 - val_loss: 0.0286 - val_accuracy: 0.2374 - val_precision: 0.7821 - val_root_mean_squared_error: 0.1691\n",
      "Epoch 199/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4173 - precision: 0.9963 - root_mean_squared_error: 0.1476 - val_loss: 0.0285 - val_accuracy: 0.2405 - val_precision: 0.7848 - val_root_mean_squared_error: 0.1690\n",
      "Epoch 200/200\n",
      "446/446 [==============================] - 11s 25ms/step - loss: 0.0218 - accuracy: 0.4144 - precision: 0.9965 - root_mean_squared_error: 0.1476 - val_loss: 0.0286 - val_accuracy: 0.2405 - val_precision: 0.7766 - val_root_mean_squared_error: 0.1690\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "from functs import myPrint\n",
    "from keras.callbacks import CSVLogger\n",
    "\n",
    "# RNN model implementation using Keras Library\n",
    "# Inspired by https://www.youtube.com/watch?v=BSpXCRTOLJA\n",
    "\n",
    "# Create sequential model object\n",
    "model = Sequential()\n",
    "\n",
    "# First input layer using LSTM with window size 64, train_x shape, & RelU activation [131 nodes]\n",
    "model.add(LSTM(num_cols, input_shape = (X_train.shape[1:]),\n",
    "               activation = 'relu', \n",
    "               return_sequences = True))\n",
    "model.add(Dropout(0.2))\n",
    "\n",
    "# First hidden layer using LSTM with RelU activation [512 nodes]\n",
    "model.add(LSTM(int(512), \n",
    "               activation = 'relu'))\n",
    "model.add(Dropout(0.2))\n",
    "\n",
    "# Final output layer using softmax activation [131 nodes]\n",
    "model.add(Dense(int(y_train.shape[1:][0]), \n",
    "                activation = 'softmax'))\n",
    "\n",
    "# optimizer used for compilation, 'Adam' refers to gradient descent method that \"adapts\" using first and second order moments\n",
    "opt = tf.keras.optimizers.Adam(learning_rate = 1e-3, \n",
    "                               decay = 1e-5)\n",
    "\n",
    "# Final compilation\n",
    "# Metrics used: accuracy, precision, RMSE\n",
    "model.compile(loss = 'mean_squared_error', \n",
    "              optimizer = opt, \n",
    "              metrics = ['accuracy', \n",
    "                         tf.keras.metrics.Precision(), \n",
    "                         tf.keras.metrics.RootMeanSquaredError()])\n",
    "\n",
    "# Model Fitting process, this is the last step\n",
    "csv_logger = CSVLogger('log.txt', \n",
    "                       append = True, \n",
    "                       separator = ';')\n",
    "history = model.fit(X_train, \n",
    "                    y_train, \n",
    "                    epochs = 200, \n",
    "                    validation_data = (X_test, \n",
    "                                       y_test), \n",
    "                    callbacks = [csv_logger])\n",
    "\n",
    "# Plot Measurment Results\n",
    "# From: https://stackoverflow.com/questions/41908379/keras-plot-training-validation-and-test-set-accuracy\n",
    "plt.plot(history.history['accuracy'])\n",
    "plt.plot(history.history['val_accuracy'])\n",
    "plt.title('model accuracy')\n",
    "plt.ylabel('accuracy')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'val'], loc='upper left')\n",
    "fig1 = plt.gcf()\n",
    "plt.show()\n",
    "fig1.savefig('Accuracy_Over_Time.png')\n",
    "\n",
    "plt.plot(history.history['loss'])\n",
    "plt.plot(history.history['val_loss'])\n",
    "plt.title('model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'val'], loc='upper left')\n",
    "fig2 = plt.gcf()\n",
    "plt.show()\n",
    "fig2.savefig('Loss_Over_Time.png')\n",
    "\n",
    "# viewable summary\n",
    "summary = model.summary(print_fn = myPrint)\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "base",
   "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": "3.9.13"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "bcb96da45b2fc45e59405d102e32af3d42527a73937a6435f97ad1b01889c6a1"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}