{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ceedc8f9",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from tensorflow.keras.models import Sequential, load_model\n",
    "from tensorflow.keras.layers import LSTM, Bidirectional, Dense, Dropout\n",
    "from tensorflow.keras.callbacks import ReduceLROnPlateau, EarlyStopping\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from sklearn.metrics import classification_report, confusion_matrix"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "98ef0040",
   "metadata": {},
   "source": [
    "## 1. Data Loading and Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b85b2bfc",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Dataset path\n",
    "PATH = \"UCI HAR Dataset/\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "42787451",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load training and test data\n",
    "X_train = np.loadtxt(PATH + \"train/X_train.txt\")\n",
    "X_test = np.loadtxt(PATH + \"test/X_test.txt\")\n",
    "y_train = np.loadtxt(PATH + \"train/y_train.txt\")\n",
    "y_test = np.loadtxt(PATH + \"test/y_test.txt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ec4ab34f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Reshape for LSTM (batch_size, timesteps, features)\n",
    "X_train = X_train.reshape(X_train.shape[0], 1, X_train.shape[1])\n",
    "X_test = X_test.reshape(X_test.shape[0], 1, X_test.shape[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9661f547",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Label encoding\n",
    "encoder = LabelEncoder()\n",
    "y_train = encoder.fit_transform(y_train)\n",
    "y_test = encoder.transform(y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ae516913",
   "metadata": {},
   "source": [
    "## 2. LSTM Model Creation and Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a02036f5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Model creation\n",
    "model = Sequential([\n",
    "    Bidirectional(LSTM(128, return_sequences=True, activation=\"tanh\"), input_shape=(X_train.shape[1], X_train.shape[2])),\n",
    "    Dropout(0.2),\n",
    "    Bidirectional(LSTM(64, return_sequences=False, activation=\"tanh\")),\n",
    "    Dropout(0.2),\n",
    "    Dense(64, activation=\"relu\"),\n",
    "    Dropout(0.1),\n",
    "    Dense(len(np.unique(y_train)), activation=\"softmax\")\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c1eae86c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Model compilation\n",
    "optimizer = tf.keras.optimizers.Adam(learning_rate=0.001)\n",
    "model.compile(loss=\"sparse_categorical_crossentropy\", optimizer=optimizer, metrics=[\"accuracy\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5421c2a2-c39f-400d-a06e-50617c27b1b1",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "21b40e87",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Callbacks to adjust learning rate and stop training early\n",
    "reduce_lr = ReduceLROnPlateau(monitor=\"val_loss\", factor=0.5, patience=5, min_lr=0.00001, verbose=1)\n",
    "early_stop = EarlyStopping(monitor=\"val_loss\", patience=10, restore_best_weights=True, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5bc38cd5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Model training\n",
    "history = model.fit(X_train, y_train, epochs=100, batch_size=128, validation_data=(X_test, y_test), callbacks=[reduce_lr, early_stop])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "164393aa",
   "metadata": {},
   "source": [
    "## 3. Model Evaluation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "82798c6b",
   "metadata": {},
   "outputs": [],
   "source": [
    "loss, accuracy = model.evaluate(X_test, y_test)\n",
    "print(f\"Test Accuracy: {accuracy:.4f}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "971d6071",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Save the model\n",
    "model.save(\"optimized_lstm_human_activity.h5\")\n",
    "print(\"✅ Optimized model successfully saved!\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6bf569ae",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Predictions on test set\n",
    "predictions = model.predict(X_test)\n",
    "predicted_labels = np.argmax(predictions, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "41a82991",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Classification report\n",
    "class_labels = [str(label) for label in np.unique(y_test)]\n",
    "print(\"\\n🔹 Classification Report:\")\n",
    "print(classification_report(y_test, predicted_labels, target_names=class_labels))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2a89410f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Confusion matrix\n",
    "cm = confusion_matrix(y_test, predicted_labels)\n",
    "plt.figure(figsize=(8,6))\n",
    "sns.heatmap(cm, annot=True, fmt=\"d\", cmap=\"Blues\", xticklabels=class_labels, yticklabels=class_labels)\n",
    "plt.xlabel(\"Predicted Class\")\n",
    "plt.ylabel(\"True Class\")\n",
    "plt.title(\"Confusion Matrix - Optimized Model\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "edc4005a",
   "metadata": {},
   "source": [
    "## 4. Validation on Synthetic Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7120363c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load the saved model\n",
    "model = load_model(\"optimized_lstm_human_activity.h5\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0687a5ac",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Generate synthetic data based on mean and standard deviation of each class\n",
    "samples_per_class = 50\n",
    "generated_data = []\n",
    "generated_labels = []"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0b8f027e",
   "metadata": {},
   "outputs": [],
   "source": [
    "for activity in range(1, 7):  # Classes from 1 to 6\n",
    "    class_data = X_train[y_train == activity]\n",
    "    feature_mean = np.mean(class_data, axis=0)\n",
    "    feature_std = np.std(class_data, axis=0)\n",
    "    \n",
    "    for _ in range(samples_per_class):\n",
    "        synthetic_sample = np.random.normal(loc=feature_mean, scale=feature_std)\n",
    "        generated_data.append(synthetic_sample)\n",
    "        generated_labels.append(activity)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "695de361",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Convert to array\n",
    "generated_data = np.array(generated_data).reshape(len(generated_data), 1, -1)\n",
    "generated_labels = np.array(generated_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0f4a73d0",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Predictions\n",
    "predictions = model.predict(generated_data)\n",
    "predicted_labels = np.argmax(predictions, axis=1) + 1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fc8e71c0",
   "metadata": {},
   "source": [
    "## 5. Visualization of Results on Synthetic Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "aec02a8d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Confusion matrix\n",
    "cm = confusion_matrix(generated_labels, predicted_labels)\n",
    "plt.figure(figsize=(8,6))\n",
    "sns.heatmap(cm, annot=True, fmt=\"d\", cmap=\"Blues\", xticklabels=range(1,7), yticklabels=range(1,7))\n",
    "plt.xlabel(\"Predicted Class\")\n",
    "plt.ylabel(\"True Class\")\n",
    "plt.title(\"Confusion Matrix - Synthetic Data\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "97b59f79",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Classification report\n",
    "print(\"\\n🔹 Classification Report:\")\n",
    "print(classification_report(generated_labels, predicted_labels))"
   ]
  }
 ],
 "metadata": {
  "jupytext": {
   "cell_metadata_filter": "-all",
   "main_language": "python",
   "notebook_metadata_filter": "-all"
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.12.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}