ModelTrainingv2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Import Libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting tensorflow_decision_forests\n",
      "  Using cached tensorflow_decision_forests-1.8.1.tar.gz (15.2 MB)\n",
      "  Preparing metadata (setup.py): started\n",
      "  Preparing metadata (setup.py): finished with status 'done'\n",
      "Requirement already satisfied: numpy in c:\\users\\sa\\appdata\\local\\programs\\python\\python312\\lib\\site-packages (from tensorflow_decision_forests) (1.26.1)\n",
      "Requirement already satisfied: pandas in c:\\users\\sa\\appdata\\local\\programs\\python\\python312\\lib\\site-packages (from tensorflow_decision_forests) (2.2.2)\n",
      "INFO: pip is looking at multiple versions of tensorflow-decision-forests to determine which version is compatible with other requirements. This could take a while.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "DEPRECATION: Loading egg at c:\\users\\sa\\appdata\\local\\programs\\python\\python312\\lib\\site-packages\\yapf-0.40.2-py3.12.egg is deprecated. pip 25.1 will enforce this behaviour change. A possible replacement is to use pip for package installation. Discussion can be found at https://github.com/pypa/pip/issues/12330\n",
      "ERROR: Could not find a version that satisfies the requirement tensorflow~=2.15.0 (from tensorflow-decision-forests) (from versions: 2.16.0rc0, 2.16.1, 2.16.2, 2.17.0rc0, 2.17.0rc1, 2.17.0, 2.17.1, 2.18.0rc0, 2.18.0rc1, 2.18.0rc2, 2.18.0)\n",
      "ERROR: No matching distribution found for tensorflow~=2.15.0\n"
     ]
    }
   ],
   "source": [
    "!pip install -U tensorflow_decision_forests\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import pandas as pd\n",
    "import tensorflow as tf\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from sklearn.metrics import accuracy_score, classification_report, confusion_matrix, precision_score, recall_score, f1_score\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Loading and Preprocessing\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Label Counts:\n",
      " label\n",
      "Yes           1286\n",
      "No             973\n",
      "I_Love_You     896\n",
      "Hello          895\n",
      "f              800\n",
      "c              797\n",
      "e              700\n",
      "Now            699\n",
      "Awkward        694\n",
      "Sorry          673\n",
      "Bathroom       672\n",
      "d              667\n",
      "Deaf           665\n",
      "Goodbye        664\n",
      "a              662\n",
      "b              653\n",
      "Name: count, dtype: int64\n",
      "Number of instances of label 'hellow': 0\n"
     ]
    }
   ],
   "source": [
    "# Load and Concatenate CSV Files\n",
    "def load_csv_files(folder_path):\n",
    "    \"\"\"\n",
    "    Load all CSV files from the specified folder and concatenate them into a single DataFrame.\n",
    "    Each file's name (without extension) is used as the label.\n",
    "    \"\"\"\n",
    "    dataframes = []\n",
    "    for file_name in os.listdir(folder_path):\n",
    "        if file_name.endswith(\".csv\"):\n",
    "            label = file_name.split(\".\")[0]  # Use the file name (without extension) as the label\n",
    "            df = pd.read_csv(os.path.join(folder_path, file_name))\n",
    "            df['label'] = label  # Add a label column\n",
    "            dataframes.append(df)\n",
    "    return pd.concat(dataframes, ignore_index=True)\n",
    "\n",
    "# Data Loading\n",
    "data_folder = r\"data\"\n",
    "df = load_csv_files(data_folder)\n",
    "\n",
    "# Select relevant columns for features and labels\n",
    "selected_columns = [\"Flex1\", \"Flex2\", \"Flex3\", \"Flex4\", \"Flex5\", \n",
    "                    \"AccelX\", \"AccelY\", \"AccelZ\", \"GyroX\", \"GyroY\", \"GyroZ\", \"label\"]\n",
    "\n",
    "# Ensure selected columns exist\n",
    "df = df[selected_columns]\n",
    "\n",
    "# Handle missing values by dropping rows with NaN values\n",
    "df = df.dropna()\n",
    "\n",
    "# Display label counts and explore the 'hellow' label\n",
    "label_counts = df[\"label\"].value_counts()\n",
    "print(\"Label Counts:\\n\", label_counts)\n",
    "\n",
    "label_hellow_count = df[df['label'] == 'hellow'].shape[0]\n",
    "print(f\"Number of instances of label 'hellow': {label_hellow_count}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Splitting: Features and Labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Label Counts:\n",
      " label\n",
      "Yes           1286\n",
      "No             973\n",
      "I_Love_You     896\n",
      "Hello          895\n",
      "f              800\n",
      "c              797\n",
      "e              700\n",
      "Now            699\n",
      "Awkward        694\n",
      "Sorry          673\n",
      "Bathroom       672\n",
      "d              667\n",
      "Deaf           665\n",
      "Goodbye        664\n",
      "a              662\n",
      "b              653\n",
      "Name: count, dtype: int64\n",
      "Number of instances of label 'hellow': 0\n"
     ]
    }
   ],
   "source": [
    "\n",
    "# Select relevant columns for features and labels\n",
    "selected_columns = [\"Flex1\", \"Flex2\", \"Flex3\", \"Flex4\", \"Flex5\", \n",
    "                    \"AccelX\", \"AccelY\", \"AccelZ\", \"GyroX\", \"GyroY\", \"GyroZ\", \"label\"]\n",
    "\n",
    "# Ensure selected columns exist\n",
    "df = df[selected_columns]\n",
    "\n",
    "# Handle missing values by dropping rows with NaN values\n",
    "df = df.dropna()\n",
    "\n",
    "# Display label counts and explore the 'hellow' label\n",
    "label_counts = df[\"label\"].value_counts()\n",
    "print(\"Label Counts:\\n\", label_counts)\n",
    "\n",
    "label_hellow_count = df[df['label'] == 'hellow'].shape[0]\n",
    "print(f\"Number of instances of label 'hellow': {label_hellow_count}\")\n",
    "\n",
    "# Data Splitting: Features and Labels\n",
    "X = df[[\"Flex1\", \"Flex2\", \"Flex3\", \"Flex4\", \"Flex5\", \"AccelX\", \"AccelY\", \"AccelZ\", \"GyroX\", \"GyroY\", \"GyroZ\"]]\n",
    "y = df[\"label\"]\n",
    "\n",
    "# Encode labels (y)\n",
    "label_encoder = LabelEncoder()\n",
    "y_encoded = label_encoder.fit_transform(y)\n",
    "\n",
    "# Split into train and test sets\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y_encoded, test_size=0.2, random_state=42, stratify=y_encoded)\n",
    "\n",
    "# Normalize data (important for neural networks)\n",
    "X_train = X_train / X_train.max()\n",
    "X_test = X_test / X_test.max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Label to Number Mapping:\n",
      "Awkward: 0\n",
      "Bathroom: 1\n",
      "Deaf: 2\n",
      "Goodbye: 3\n",
      "Hello: 4\n",
      "I_Love_You: 5\n",
      "No: 6\n",
      "Now: 7\n",
      "Sorry: 8\n",
      "Yes: 9\n",
      "a: 10\n",
      "b: 11\n",
      "c: 12\n",
      "d: 13\n",
      "e: 14\n",
      "f: 15\n"
     ]
    }
   ],
   "source": [
    "# Create the mapping between labels and their numeric encoded values\n",
    "label_mapping = dict(zip(label_encoder.classes_, range(len(label_encoder.classes_))))\n",
    "\n",
    "# Print the label to number mapping\n",
    "print(\"Label to Number Mapping:\")\n",
    "for label, number in label_mapping.items():\n",
    "    print(f\"{label}: {number}\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Build a simple neural network model in TensorFlow"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/50\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\sa\\AppData\\Local\\Programs\\Python\\Python312\\Lib\\site-packages\\keras\\src\\layers\\core\\dense.py:87: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.\n",
      "  super().__init__(activity_regularizer=activity_regularizer, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.1678 - loss: 2.5867 - val_accuracy: 0.4480 - val_loss: 1.7218 - learning_rate: 0.0010\n",
      "Epoch 2/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.4719 - loss: 1.6822 - val_accuracy: 0.6798 - val_loss: 1.1595 - learning_rate: 0.0010\n",
      "Epoch 3/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - accuracy: 0.5850 - loss: 1.2947 - val_accuracy: 0.6835 - val_loss: 1.0179 - learning_rate: 0.0010\n",
      "Epoch 4/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.6522 - loss: 1.0775 - val_accuracy: 0.7270 - val_loss: 0.9024 - learning_rate: 0.0010\n",
      "Epoch 5/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.6874 - loss: 0.9525 - val_accuracy: 0.7774 - val_loss: 0.7472 - learning_rate: 0.0010\n",
      "Epoch 6/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7252 - loss: 0.8693 - val_accuracy: 0.8000 - val_loss: 0.6975 - learning_rate: 0.0010\n",
      "Epoch 7/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 1ms/step - accuracy: 0.7479 - loss: 0.7817 - val_accuracy: 0.8246 - val_loss: 0.6136 - learning_rate: 0.0010\n",
      "Epoch 8/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7693 - loss: 0.7249 - val_accuracy: 0.8028 - val_loss: 0.6705 - learning_rate: 0.0010\n",
      "Epoch 9/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7880 - loss: 0.6400 - val_accuracy: 0.8230 - val_loss: 0.5802 - learning_rate: 0.0010\n",
      "Epoch 10/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - accuracy: 0.7963 - loss: 0.6307 - val_accuracy: 0.8327 - val_loss: 0.5313 - learning_rate: 0.0010\n",
      "Epoch 11/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8051 - loss: 0.5961 - val_accuracy: 0.8270 - val_loss: 0.5363 - learning_rate: 0.0010\n",
      "Epoch 12/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - accuracy: 0.8038 - loss: 0.6123 - val_accuracy: 0.8355 - val_loss: 0.4919 - learning_rate: 0.0010\n",
      "Epoch 13/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - accuracy: 0.8180 - loss: 0.5465 - val_accuracy: 0.8504 - val_loss: 0.4730 - learning_rate: 0.0010\n",
      "Epoch 14/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8246 - loss: 0.5340 - val_accuracy: 0.8472 - val_loss: 0.4717 - learning_rate: 0.0010\n",
      "Epoch 15/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8377 - loss: 0.5120 - val_accuracy: 0.8617 - val_loss: 0.4593 - learning_rate: 0.0010\n",
      "Epoch 16/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8353 - loss: 0.5121 - val_accuracy: 0.8746 - val_loss: 0.4195 - learning_rate: 0.0010\n",
      "Epoch 17/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - accuracy: 0.8378 - loss: 0.4989 - val_accuracy: 0.8786 - val_loss: 0.4037 - learning_rate: 0.0010\n",
      "Epoch 18/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - accuracy: 0.8456 - loss: 0.4916 - val_accuracy: 0.8766 - val_loss: 0.4031 - learning_rate: 0.0010\n",
      "Epoch 19/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8506 - loss: 0.4711 - val_accuracy: 0.8798 - val_loss: 0.4057 - learning_rate: 0.0010\n",
      "Epoch 20/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - accuracy: 0.8543 - loss: 0.4588 - val_accuracy: 0.8806 - val_loss: 0.3720 - learning_rate: 0.0010\n",
      "Epoch 21/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8516 - loss: 0.4490 - val_accuracy: 0.8839 - val_loss: 0.3710 - learning_rate: 0.0010\n",
      "Epoch 22/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8454 - loss: 0.4630 - val_accuracy: 0.8911 - val_loss: 0.3590 - learning_rate: 0.0010\n",
      "Epoch 23/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8507 - loss: 0.4392 - val_accuracy: 0.8843 - val_loss: 0.3527 - learning_rate: 0.0010\n",
      "Epoch 24/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - accuracy: 0.8638 - loss: 0.4173 - val_accuracy: 0.8875 - val_loss: 0.3680 - learning_rate: 0.0010\n",
      "Epoch 25/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8674 - loss: 0.4105 - val_accuracy: 0.8859 - val_loss: 0.3562 - learning_rate: 0.0010\n",
      "Epoch 26/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - accuracy: 0.8676 - loss: 0.3974 - val_accuracy: 0.8931 - val_loss: 0.3409 - learning_rate: 0.0010\n",
      "Epoch 27/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8712 - loss: 0.4110 - val_accuracy: 0.8871 - val_loss: 0.3594 - learning_rate: 0.0010\n",
      "Epoch 28/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8700 - loss: 0.4041 - val_accuracy: 0.8774 - val_loss: 0.3590 - learning_rate: 0.0010\n",
      "Epoch 29/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8713 - loss: 0.3907 - val_accuracy: 0.9004 - val_loss: 0.3443 - learning_rate: 0.0010\n",
      "Epoch 30/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8777 - loss: 0.3707 - val_accuracy: 0.9012 - val_loss: 0.3297 - learning_rate: 1.0000e-04\n",
      "Epoch 31/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8924 - loss: 0.3533 - val_accuracy: 0.8996 - val_loss: 0.3284 - learning_rate: 1.0000e-04\n",
      "Epoch 32/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8916 - loss: 0.3445 - val_accuracy: 0.9032 - val_loss: 0.3264 - learning_rate: 1.0000e-04\n",
      "Epoch 33/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - accuracy: 0.8950 - loss: 0.3386 - val_accuracy: 0.9036 - val_loss: 0.3233 - learning_rate: 1.0000e-04\n",
      "Epoch 34/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8887 - loss: 0.3571 - val_accuracy: 0.9020 - val_loss: 0.3221 - learning_rate: 1.0000e-04\n",
      "Epoch 35/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8872 - loss: 0.3503 - val_accuracy: 0.9016 - val_loss: 0.3203 - learning_rate: 1.0000e-04\n",
      "Epoch 36/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8874 - loss: 0.3412 - val_accuracy: 0.9020 - val_loss: 0.3226 - learning_rate: 1.0000e-04\n",
      "Epoch 37/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8941 - loss: 0.3311 - val_accuracy: 0.9056 - val_loss: 0.3197 - learning_rate: 1.0000e-04\n",
      "Epoch 38/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8844 - loss: 0.3573 - val_accuracy: 0.9036 - val_loss: 0.3194 - learning_rate: 1.0000e-04\n",
      "Epoch 39/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - accuracy: 0.8891 - loss: 0.3453 - val_accuracy: 0.9044 - val_loss: 0.3176 - learning_rate: 1.0000e-04\n",
      "Epoch 40/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8889 - loss: 0.3422 - val_accuracy: 0.9004 - val_loss: 0.3202 - learning_rate: 1.0000e-04\n",
      "Epoch 41/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - accuracy: 0.8926 - loss: 0.3332 - val_accuracy: 0.9028 - val_loss: 0.3179 - learning_rate: 1.0000e-04\n",
      "Epoch 42/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.8945 - loss: 0.3409 - val_accuracy: 0.9052 - val_loss: 0.3149 - learning_rate: 1.0000e-04\n",
      "Epoch 43/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8951 - loss: 0.3323 - val_accuracy: 0.9040 - val_loss: 0.3170 - learning_rate: 1.0000e-04\n",
      "Epoch 44/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8869 - loss: 0.3582 - val_accuracy: 0.9052 - val_loss: 0.3140 - learning_rate: 1.0000e-04\n",
      "Epoch 45/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8930 - loss: 0.3250 - val_accuracy: 0.9048 - val_loss: 0.3173 - learning_rate: 1.0000e-04\n",
      "Epoch 46/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8904 - loss: 0.3324 - val_accuracy: 0.9040 - val_loss: 0.3147 - learning_rate: 1.0000e-04\n",
      "Epoch 47/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - accuracy: 0.8920 - loss: 0.3418 - val_accuracy: 0.9060 - val_loss: 0.3119 - learning_rate: 1.0000e-04\n",
      "Epoch 48/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8948 - loss: 0.3431 - val_accuracy: 0.9073 - val_loss: 0.3134 - learning_rate: 1.0000e-04\n",
      "Epoch 49/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8847 - loss: 0.3587 - val_accuracy: 0.9069 - val_loss: 0.3102 - learning_rate: 1.0000e-04\n",
      "Epoch 50/50\n",
      "\u001b[1m310/310\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8911 - loss: 0.3409 - val_accuracy: 0.9065 - val_loss: 0.3145 - learning_rate: 1.0000e-04\n"
     ]
    }
   ],
   "source": [
    "# Build a deeper model with dropout and L2 regularization\n",
    "model = tf.keras.Sequential([\n",
    "    tf.keras.layers.Dense(128, activation='relu', input_shape=(X_train.shape[1],)),\n",
    "    tf.keras.layers.Dropout(0.2),  # Adding dropout to reduce overfitting\n",
    "    tf.keras.layers.Dense(64, activation='relu'),\n",
    "    tf.keras.layers.Dropout(0.2),\n",
    "    tf.keras.layers.Dense(32, activation='relu'),\n",
    "    tf.keras.layers.Dense(len(label_encoder.classes_), activation='softmax')  # For multi-class classification\n",
    "])\n",
    "\n",
    "# Compile the model with Adam optimizer and categorical crossentropy loss\n",
    "model.compile(optimizer=tf.keras.optimizers.Adam(learning_rate=0.001),\n",
    "              loss='sparse_categorical_crossentropy',  # Use 'sparse_categorical_crossentropy' for integer labels\n",
    "              metrics=['accuracy'])\n",
    "\n",
    "# Implement Early Stopping and ReduceLROnPlateau\n",
    "early_stopping = tf.keras.callbacks.EarlyStopping(monitor='val_loss', patience=10, restore_best_weights=True)\n",
    "lr_scheduler = tf.keras.callbacks.ReduceLROnPlateau(monitor='val_loss', factor=0.1, patience=3)\n",
    "\n",
    "# Train the model\n",
    "history = model.fit(X_train, y_train, epochs=50, batch_size=32, \n",
    "                    validation_data=(X_test, y_test),\n",
    "                    callbacks=[early_stopping, lr_scheduler])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m78/78\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9007 - loss: 0.3022\n",
      "Test Accuracy: 0.9068548679351807\n",
      "\u001b[1m78/78\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 2ms/step\n",
      "Model Accuracy: 0.9068548387096774\n"
     ]
    }
   ],
   "source": [
    "# Evaluate the model on test data\n",
    "loss, accuracy = model.evaluate(X_test, y_test)\n",
    "print(f\"Test Accuracy: {accuracy}\")\n",
    "\n",
    "# Model Evaluation: Accuracy and Classification Report\n",
    "y_pred = model.predict(X_test)\n",
    "y_pred_classes = y_pred.argmax(axis=-1)\n",
    "\n",
    "# Model Accuracy\n",
    "accuracy = accuracy_score(y_test, y_pred_classes)\n",
    "print(f\"Model Accuracy: {accuracy}\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model Evaluation: Accuracy and Classification Report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Classification Report:\n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.99      0.78      0.88       139\n",
      "           1       0.92      0.91      0.92       134\n",
      "           2       0.96      0.99      0.98       133\n",
      "           3       1.00      0.94      0.97       133\n",
      "           4       0.74      0.94      0.83       179\n",
      "           5       0.95      0.84      0.89       179\n",
      "           6       0.76      0.84      0.80       195\n",
      "           7       1.00      0.89      0.94       140\n",
      "           8       0.99      0.94      0.97       135\n",
      "           9       0.80      0.86      0.83       257\n",
      "          10       1.00      0.96      0.98       132\n",
      "          11       1.00      0.99      1.00       131\n",
      "          12       0.88      0.84      0.86       160\n",
      "          13       0.99      0.96      0.98       133\n",
      "          14       0.99      0.94      0.96       140\n",
      "          15       0.89      0.97      0.93       160\n",
      "\n",
      "    accuracy                           0.91      2480\n",
      "   macro avg       0.93      0.91      0.92      2480\n",
      "weighted avg       0.92      0.91      0.91      2480\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Precision: 0.9152958103782508\n",
      "Recall: 0.9068548387096774\n",
      "F1 Score: 0.9084704961656231\n"
     ]
    }
   ],
   "source": [
    "# Classification Report\n",
    "print(\"\\nClassification Report:\\n\", classification_report(y_test, y_pred_classes))\n",
    "\n",
    "# Confusion Matrix\n",
    "conf_matrix = confusion_matrix(y_test, y_pred_classes)\n",
    "plt.figure(figsize=(10, 6))\n",
    "sns.heatmap(conf_matrix, annot=True, fmt=\"d\", cmap=\"Blues\", xticklabels=label_encoder.classes_, yticklabels=label_encoder.classes_)\n",
    "plt.title(\"Confusion Matrix\")\n",
    "plt.xlabel(\"Predicted\")\n",
    "plt.ylabel(\"Actual\")\n",
    "plt.show()\n",
    "\n",
    "# Precision, Recall, F1-Score\n",
    "precision = precision_score(y_test, y_pred_classes, average=\"weighted\")\n",
    "recall = recall_score(y_test, y_pred_classes, average=\"weighted\")\n",
    "f1 = f1_score(y_test, y_pred_classes, average=\"weighted\")\n",
    "print(f\"Precision: {precision}\")\n",
    "print(f\"Recall: {recall}\")\n",
    "print(f\"F1 Score: {f1}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Save the Trained Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:You are saving your model as an HDF5 file via `model.save()` or `keras.saving.save_model(model)`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')` or `keras.saving.save_model(model, 'my_model.keras')`. \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: C:\\Users\\sa\\AppData\\Local\\Temp\\tmp5hv_zc36\\assets\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: C:\\Users\\sa\\AppData\\Local\\Temp\\tmp5hv_zc36\\assets\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saved artifact at 'C:\\Users\\sa\\AppData\\Local\\Temp\\tmp5hv_zc36'. The following endpoints are available:\n",
      "\n",
      "* Endpoint 'serve'\n",
      "  args_0 (POSITIONAL_ONLY): TensorSpec(shape=(None, 11), dtype=tf.float32, name='keras_tensor_36')\n",
      "Output Type:\n",
      "  TensorSpec(shape=(None, 16), dtype=tf.float32, name=None)\n",
      "Captures:\n",
      "  1913632717328: TensorSpec(shape=(), dtype=tf.resource, name=None)\n",
      "  1913733999184: TensorSpec(shape=(), dtype=tf.resource, name=None)\n",
      "  1913733994000: TensorSpec(shape=(), dtype=tf.resource, name=None)\n",
      "  1913733998416: TensorSpec(shape=(), dtype=tf.resource, name=None)\n",
      "  1913733997264: TensorSpec(shape=(), dtype=tf.resource, name=None)\n",
      "  1913734001296: TensorSpec(shape=(), dtype=tf.resource, name=None)\n",
      "  1913734000720: TensorSpec(shape=(), dtype=tf.resource, name=None)\n",
      "  1913733996880: TensorSpec(shape=(), dtype=tf.resource, name=None)\n",
      "TFLite model has been saved as 'model.tflite'\n"
     ]
    }
   ],
   "source": [
    "# Save the Trained Model\n",
    "# Save the model as a .h5 file\n",
    "model.save(\"improved_model.h5\")\n",
    "\n",
    "# Convert the trained model to TFLite format\n",
    "converter = tf.lite.TFLiteConverter.from_keras_model(model)\n",
    "\n",
    "# Perform the conversion\n",
    "tflite_model = converter.convert()\n",
    "\n",
    "# Save the converted model to a file\n",
    "with open(\"model.tflite\", \"wb\") as f:\n",
    "    f.write(tflite_model)\n",
    "\n",
    "print(\"TFLite model has been saved as 'model.tflite'\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.12.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}