Hw8.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "b0Yf4NBJUSNM"
   },
   "source": [
    "# Створення нейронної мережі\n",
    "\n",
    "У цьому завданні ми створимо повнозв'язну нейронну мережу, використовуючи при цьому низькорівневі механізми tensorflow.\n",
    "\n",
    "Архітектура нейромережі представлена на наступному малюнку. Як бачиш, у ній є один вхідний шар, два приховані, а також вихідний шар. В якості активаційної функції у прихованих шарах буде використовуватись сигмоїда. На вихідному шарі ми використовуємо softmax.\n",
    "\n",
    "Частина коду зі створення мережі вже написана, тобі потрібно заповнити пропуски у вказаних місцях."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "01rZWUu0USNQ"
   },
   "source": [
    "## Архітектура нейронної мережі\n",
    "\n",
    "<img src=\"http://cs231n.github.io/assets/nn1/neural_net2.jpeg\" alt=\"nn\" style=\"width: 400px;\"/>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "LLvIZ705Qw_V"
   },
   "source": [
    "## Про датасет MNIST\n",
    "\n",
    "Дану нейромережу ми будемо вивчати на датасеті MNIST. Цей датасет являє собою велику кількість зображень рукописних цифр розміром $28 \\times 28$ пікселів. Кожен піксель приймає значення від 0 до 255.\n",
    "\n",
    "Як і раніше, датасет буде розділений на навчальну та тестову вибірки. При цьому ми виконаємо нормалізацію всіх зображень, щоб значення пікселів знаходилось у проміжку від 0 до 1, розділивши яскравість кожного пікселя на 255.\n",
    "\n",
    "Окрім того, архітектура нейронної мережі очікує на вхід вектор. У нашому ж випадку кожен об'єкт вибірки являє собою матрицю. Що ж робити? У цьому завданні ми \"розтягнемо\" матрицю $28 \\times 28$, отримавши при цьому вектор, що складається з 784 елементів.\n",
    "\n",
    "![MNIST Dataset](https://www.researchgate.net/profile/Steven-Young-5/publication/306056875/figure/fig1/AS:393921575309346@1470929630835/Example-images-from-the-MNIST-dataset.png)\n",
    "\n",
    "Більше інформації про датасет можна знайти [тут](http://yann.lecun.com/exdb/mnist/)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "id": "il_0_5OyUSNR"
   },
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import random\n",
    "import keras as K\n",
    "import seaborn as sns\n",
    "from sklearn.metrics import confusion_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "id": "cd-1_abTUSNS"
   },
   "outputs": [],
   "source": [
    "num_classes = 10 # загальна кількість класів, у нашому випадку це цифри від 0 до 9\n",
    "num_features = 784 # кількість атрибутів вхідного вектора 28 * 28 = 784\n",
    "\n",
    "learning_rate = 0.001 # швидкість навчання нейронної мережі\n",
    "training_steps = 3000 # максимальне число епох\n",
    "batch_size = 256 # перераховувати ваги мережі ми будемо не на всій вибірці, а на її випадковій підмножині з batch_size елементів\n",
    "display_step = 100 # кожні 100 ітерацій ми будемо показувати поточне значення функції втрат і точності\n",
    "\n",
    "n_hidden_1 = 128 # кількість нейронів 1-го шару\n",
    "n_hidden_2 = 256 # кількість нейронів 2-го шару"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "pGTXiRyTUSNT",
    "outputId": "68f5b9d9-6c78-4bda-b26f-69aa82e93f20"
   },
   "outputs": [],
   "source": [
    "# from tensorflow.keras.datasets import mnist\n",
    "from keras.datasets import mnist\n",
    "\n",
    "# Завантажуємо датасет\n",
    "(x_train, y_train), (x_test, y_test) = mnist.load_data()\n",
    "\n",
    "# Перетворюємо цілочисельні пікселі на тип float32\n",
    "x_train, x_test = np.array(x_train, np.float32), np.array(x_test, np.float32)\n",
    "\n",
    "# Перетворюємо матриці розміром 28x28 пікселів у вектор з 784 елементів\n",
    "x_train, x_test = x_train.reshape([-1, num_features]), x_test.reshape([-1, num_features])\n",
    "\n",
    "# Нормалізуємо значення пікселів\n",
    "x_train, x_test = x_train / 255., x_test / 255.\n",
    "\n",
    "# Перемішаємо тренувальні дані\n",
    "train_data = tf.data.Dataset.from_tensor_slices((x_train, y_train))\n",
    "train_data = train_data.repeat().shuffle(5000).batch(batch_size).prefetch(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "id": "FkRmCQjnUSNV"
   },
   "outputs": [],
   "source": [
    "# Створимо нейронну мережу\n",
    "\n",
    "class DenseLayer(tf.Module):\n",
    "    def __init__(self, in_features, out_features, name=None):\n",
    "        super().__init__(name=name)\n",
    "        self.w = tf.Variable(\n",
    "            tf.random.normal([in_features, out_features]), name=\"w\"\n",
    "        )\n",
    "        self.b = tf.Variable(tf.zeros([out_features]), name=\"b\")\n",
    "\n",
    "    def __call__(self, x, activation=0):\n",
    "        y = tf.matmul(x, self.w) + self.b\n",
    "        if activation != 0:\n",
    "            return tf.nn.softmax(y)\n",
    "        else:\n",
    "            return tf.nn.sigmoid(y)\n",
    "\n",
    "class NN(tf.Module):\n",
    "  def __init__(self, name=None):\n",
    "    super().__init__(name=name)\n",
    "    # Перший шар, який складається з 128 нейронів\n",
    "    self.layer_1 = DenseLayer(in_features=num_features, out_features=n_hidden_1)\n",
    "    # Другий шар, який складається з 256 нейронів\n",
    "    self.layer_2 = DenseLayer(in_features=n_hidden_1, out_features=n_hidden_2)\n",
    "    # Вихідний шар\n",
    "    self.out_layer = DenseLayer(in_features=n_hidden_2, out_features=num_classes)\n",
    "\n",
    "\n",
    "  def __call__(self, x):\n",
    "    # Передача даних через перші два шари та вихідний шар з функцією активації softmax\n",
    "    x1 = self.layer_1(x, activation=0)\n",
    "    x2 = self.layer_2(x1, activation=0)\n",
    "    x3 = self.out_layer(x2, activation=1)\n",
    "\n",
    "\n",
    "    return x3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "id": "LIf3o7VAUSNV"
   },
   "outputs": [],
   "source": [
    "# В якості функції помилки в даному випадку зручно взяти крос-ентропію\n",
    "def cross_entropy(y_pred, y_true):\n",
    "    # Закодувати label в one hot vector\n",
    "    y_true = tf.one_hot(y_true, depth=num_classes)\n",
    "\n",
    "    # Значення передбачення, щоб уникнути помилки log(0).\n",
    "    y_pred = tf.clip_by_value(y_pred, 1e-9, 1.)\n",
    "\n",
    "    # Обчислення крос-ентропії\n",
    "    return tf.reduce_mean(-tf.reduce_sum(y_true * tf.math.log(y_pred)))\n",
    "\n",
    "# Як метрику якості використовуємо точність\n",
    "def accuracy(y_pred, y_true):\n",
    "    # Місце для вашого коду\n",
    "    y_pred_argmax = tf.argmax(y_pred, axis=1)\n",
    "    y_true_cast = tf.cast(y_true, tf.int64)\n",
    "\n",
    "    # Порівняння y_pred з y_true\n",
    "    correct_predictions = tf.equal(y_pred_argmax, y_true_cast)\n",
    "\n",
    "    # Знаходження середньої точності \n",
    "    accuracy = tf.reduce_mean(tf.cast(correct_predictions, tf.float32))\n",
    "\n",
    "    return accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "id": "MQeT1yatUSNW"
   },
   "outputs": [],
   "source": [
    "# Створимо екземпляр нейронної мережі\n",
    "neural_net = NN(name=\"mnist\")\n",
    "\n",
    "# Функція навчання нейромережі\n",
    "def train(neural_net, input_x, output_y):\n",
    "  # Для налаштування вагів мережі будемо використовувати стохастичний градієнтний спуск\n",
    "  optimizer = tf.optimizers.SGD(learning_rate)\n",
    "\n",
    "  # Активація автоматичного диференціювання\n",
    "  with tf.GradientTape() as g:\n",
    "    pred = neural_net(input_x)\n",
    "    loss = cross_entropy(pred, output_y)\n",
    "\n",
    "    # Отримаємо список оптимізованих параметрів\n",
    "    trainable_variables = neural_net.trainable_variables\n",
    "    # Обчислимо за ними значення градієнта\n",
    "    gradients = g.gradient(loss, trainable_variables)\n",
    "\n",
    "    # Модифікуємо параметри\n",
    "    optimizer.apply_gradients(zip(gradients, trainable_variables))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "id": "fnyns9lBfpQZ"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training step: 100, loss: 335.8466, accuracy: 0.6719\n",
      "Training step: 200, loss: 225.6522, accuracy: 0.7383\n",
      "Training step: 300, loss: 174.2234, accuracy: 0.8047\n",
      "Training step: 400, loss: 175.2688, accuracy: 0.8516\n",
      "Training step: 500, loss: 171.9912, accuracy: 0.8359\n",
      "Training step: 600, loss: 161.6163, accuracy: 0.8047\n",
      "Training step: 700, loss: 91.9465, accuracy: 0.9023\n",
      "Training step: 800, loss: 145.0899, accuracy: 0.8359\n",
      "Training step: 900, loss: 95.8254, accuracy: 0.9023\n",
      "Training step: 1000, loss: 95.9397, accuracy: 0.8750\n",
      "Training step: 1100, loss: 113.3781, accuracy: 0.8906\n",
      "Training step: 1200, loss: 99.8196, accuracy: 0.8828\n",
      "Training step: 1300, loss: 116.2426, accuracy: 0.8672\n",
      "Training step: 1400, loss: 95.8380, accuracy: 0.9180\n",
      "Training step: 1500, loss: 66.1934, accuracy: 0.9336\n",
      "Training step: 1600, loss: 99.3109, accuracy: 0.9062\n",
      "Training step: 1700, loss: 66.3795, accuracy: 0.8984\n",
      "Training step: 1800, loss: 98.9237, accuracy: 0.8945\n",
      "Training step: 1900, loss: 63.5466, accuracy: 0.9375\n",
      "Training step: 2000, loss: 68.5940, accuracy: 0.9180\n",
      "Training step: 2100, loss: 48.3429, accuracy: 0.9375\n",
      "Training step: 2200, loss: 61.2144, accuracy: 0.9336\n",
      "Training step: 2300, loss: 59.0930, accuracy: 0.9180\n",
      "Training step: 2400, loss: 60.5374, accuracy: 0.9492\n",
      "Training step: 2500, loss: 58.1170, accuracy: 0.9297\n",
      "Training step: 2600, loss: 52.7027, accuracy: 0.9492\n",
      "Training step: 2700, loss: 61.5948, accuracy: 0.9297\n",
      "Training step: 2800, loss: 101.1820, accuracy: 0.9023\n",
      "Training step: 2900, loss: 59.6363, accuracy: 0.9297\n",
      "Training step: 3000, loss: 63.9536, accuracy: 0.9258\n"
     ]
    }
   ],
   "source": [
    "# Тренування мережі\n",
    "\n",
    "loss_history = []  # кожні display_step кроків зберігай в цьому списку поточну помилку нейромережі\n",
    "accuracy_history = [] # кожні display_step кроків зберігай в цьому списку поточну точність нейромережі\n",
    "\n",
    "# У цьому циклі ми будемо проводити навчання нейронної мережі\n",
    "# із тренувального датасета train_data вилучи випадкову підмножину, на якій\n",
    "# відбудеться тренування. Використовуй метод take, доступний для тренувального датасета.\n",
    "for step, (batch_x, batch_y) in enumerate(train_data.take(training_steps), 1):\n",
    "    # Оновлюємо ваги нейронної мережі\n",
    "    train(neural_net, batch_x, batch_y)\n",
    "\n",
    "    if step % display_step == 0:\n",
    "        pred = neural_net(batch_x)\n",
    "        loss = cross_entropy(pred, batch_y)\n",
    "        loss_history.append(loss)\n",
    "        acc = accuracy(pred, batch_y)\n",
    "        accuracy_history.append(acc)\n",
    "        print(f\"Training step: {step}, loss: {loss:.4f}, accuracy: {acc:.4f}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "id": "_yCBfG6MbQB2"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Виведіть графіки залежності зміни точності і втрат від кроку\n",
    "# Якщо все зроблено правильно, то точність повинна зростати, а втрати зменшуватись\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Виведіть графік функції втрат\n",
    "plt.figure(figsize=(12, 8))\n",
    "plt.plot(loss_history, label='Loss', color=\"teal\")\n",
    "plt.title('Dependence of Loss on Training Step')\n",
    "plt.xlabel('Training Step')\n",
    "plt.ylabel('Loss')\n",
    "plt.grid(True)\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "# Виведіть графік точності\n",
    "plt.figure(figsize=(12, 8))\n",
    "plt.plot(accuracy_history, label='Accuracy', color=\"slateblue\")\n",
    "plt.title('Dependence of Accuracy on Training Step')\n",
    "plt.xlabel('Training Step')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.grid(True)\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "id": "LE3g4gDyUSNY"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy on Training data: 0.9265000224113464\n",
      "Accuracy on Test data: 0.9023000001907349\n"
     ]
    }
   ],
   "source": [
    "# Обчисліть точність навченої нейромережі\n",
    "pred_train = neural_net(x_train)\n",
    "accuracy_train_data = accuracy(pred_train, y_train)\n",
    "print(f\"Accuracy on Training data: {accuracy_train_data}\")\n",
    "\n",
    "# Тестування моделі на тестових даних\n",
    "pred_test = neural_net(x_test)\n",
    "accuracy_test_data = accuracy(pred_test, y_test)\n",
    "print(f\"Accuracy on Test data: {accuracy_test_data}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "id": "_EEHAubOUSNY"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x500 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Протестуйте навчену нейромережу на 10 зображеннях. З тестової вибірки візьміть 5\n",
    "# випадкових зображень і передайте їх у нейронну мережу.\n",
    "random_indices = np.random.choice(len(x_test), 5)\n",
    "images_to_test = x_test[random_indices]\n",
    "labels_to_test = y_test[random_indices]\n",
    "prediction = neural_net(images_to_test)\n",
    "\n",
    "# Виведіть зображення та випишіть  поруч відповідь нейромережі.\n",
    "plt.figure(figsize=(15, 5))\n",
    "for elem in range(5):\n",
    "    plt.subplot(1, 5, elem+1)\n",
    "    plt.imshow(images_to_test[elem].reshape((28, 28)), cmap='gray')\n",
    "    plt.title(f\"Prediction: {tf.argmax(prediction[elem], axis=0).numpy()}, Actual: {labels_to_test[elem]}\")\n",
    "    plt.axis('off')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "####  Зробіть висновок про те, чи помиляється твоя нейронна мережа, і якщо так, то як часто?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Нейронна мережа, навчена на базі даних MNIST, досягла досить гарної точності приблизно 90,23%. Це вказує на те, що модель добре класифікує рукописні цифри, але все таки має невеликий відсоток помилок"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "provenance": []
  },
  "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": "3.10.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}