Optimizersr_LossValues_Accuracies_Precisions_Recalls_F-scores..ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "99276333",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 17000 images belonging to 10 classes.\n",
      "Found 2999 images belonging to 10 classes.\n",
      "Epoch 1/10\n",
      "266/266 [==============================] - 17s 61ms/step - loss: 1.1045 - accuracy: 0.6404 - val_loss: 0.2834 - val_accuracy: 0.9183\n",
      "Epoch 2/10\n",
      "266/266 [==============================] - 16s 61ms/step - loss: 0.5400 - accuracy: 0.8311 - val_loss: 0.1665 - val_accuracy: 0.9463\n",
      "Epoch 3/10\n",
      "266/266 [==============================] - 16s 59ms/step - loss: 0.4022 - accuracy: 0.8720 - val_loss: 0.1609 - val_accuracy: 0.9480\n",
      "Epoch 4/10\n",
      "266/266 [==============================] - 16s 60ms/step - loss: 0.3276 - accuracy: 0.8976 - val_loss: 0.1185 - val_accuracy: 0.9670\n",
      "Epoch 5/10\n",
      "266/266 [==============================] - 16s 59ms/step - loss: 0.2830 - accuracy: 0.9112 - val_loss: 0.0824 - val_accuracy: 0.9747\n",
      "Epoch 6/10\n",
      "266/266 [==============================] - 16s 60ms/step - loss: 0.2595 - accuracy: 0.9185 - val_loss: 0.0776 - val_accuracy: 0.9760\n",
      "Epoch 7/10\n",
      "266/266 [==============================] - 16s 60ms/step - loss: 0.2316 - accuracy: 0.9273 - val_loss: 0.0985 - val_accuracy: 0.9707\n",
      "Epoch 8/10\n",
      "266/266 [==============================] - 16s 60ms/step - loss: 0.2157 - accuracy: 0.9333 - val_loss: 0.0624 - val_accuracy: 0.9797\n",
      "Epoch 9/10\n",
      "266/266 [==============================] - 16s 60ms/step - loss: 0.2055 - accuracy: 0.9346 - val_loss: 0.0613 - val_accuracy: 0.9787\n",
      "Epoch 10/10\n",
      "266/266 [==============================] - 16s 62ms/step - loss: 0.1866 - accuracy: 0.9418 - val_loss: 0.0649 - val_accuracy: 0.9770\n",
      "47/47 - 1s - loss: 0.0649 - accuracy: 0.9770 - 1s/epoch - 29ms/step\n",
      "Test accuracy: 0.9769923090934753\n"
     ]
    }
   ],
   "source": [
    "# import necessary libraries\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras import layers\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "\n",
    "# define the paths to the train and test folders on your desktop\n",
    "train_path = 'C:/Users/chand/Desktop/New folder/DeepLearning/Dataset/train'\n",
    "test_path = 'C:/Users/chand/Desktop/New folder/DeepLearning/Dataset/test'\n",
    "\n",
    "# define image and batch size\n",
    "img_height, img_width = 28, 28\n",
    "batch_size = 64\n",
    "\n",
    "# create ImageDataGenerator objects for data augmentation\n",
    "train_datagen = ImageDataGenerator(\n",
    "    rescale=1.0/255.0,\n",
    "    rotation_range=20,\n",
    "    zoom_range=0.2,\n",
    "    shear_range=0.2,\n",
    "    width_shift_range=0.2,\n",
    "    height_shift_range=0.2,\n",
    "    horizontal_flip=False,\n",
    "    vertical_flip=False,\n",
    "    fill_mode='nearest'\n",
    ")\n",
    "\n",
    "test_datagen = ImageDataGenerator(\n",
    "    rescale=1.0/255.0\n",
    ")\n",
    "\n",
    "# create flow_from_directory generators for train and test datasets\n",
    "train_generator = train_datagen.flow_from_directory(\n",
    "    directory=train_path,\n",
    "    target_size=(img_height, img_width),\n",
    "    color_mode=\"grayscale\",\n",
    "    batch_size=batch_size,\n",
    "    class_mode=\"categorical\",\n",
    "    shuffle=True,\n",
    "    seed=42\n",
    ")\n",
    "\n",
    "test_generator = test_datagen.flow_from_directory(\n",
    "    directory=test_path,\n",
    "    target_size=(img_height, img_width),\n",
    "    color_mode=\"grayscale\",\n",
    "    batch_size=batch_size,\n",
    "    class_mode=\"categorical\",\n",
    "    shuffle=False,\n",
    "    seed=42\n",
    ")\n",
    "\n",
    "# define the model architecture\n",
    "model = keras.Sequential(\n",
    "    [\n",
    "        keras.Input(shape=(img_height, img_width, 1)),\n",
    "        layers.Conv2D(32, kernel_size=(3, 3), activation=\"relu\"),\n",
    "        layers.MaxPooling2D(pool_size=(2, 2)),\n",
    "        layers.Flatten(),\n",
    "        layers.Dense(128, activation=\"relu\"),\n",
    "        layers.Dense(train_generator.num_classes, activation=\"softmax\"),\n",
    "    ]\n",
    ")\n",
    "\n",
    "# compile the model\n",
    "model.compile(optimizer=\"adam\", loss=\"categorical_crossentropy\", metrics=[\"accuracy\"])\n",
    "\n",
    "# train the model\n",
    "# MODEL.summary\n",
    "\n",
    "history = model.fit(train_generator, epochs=10, validation_data=test_generator)\n",
    "\n",
    "# evaluate the model on test set\n",
    "test_loss, test_acc = model.evaluate(test_generator, verbose=2)\n",
    "print(\"Test accuracy:\", test_acc)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "b1cdf4ad",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "47/47 [==============================] - 3s 29ms/step\n",
      "Optimizer: SGD\n",
      "Test loss: 0.3211318254470825\n",
      "Test accuracy: 0.9033011198043823\n",
      "Precision: 0.99776953\n",
      "Recall: 0.99407405\n",
      "F-score: 0.9959184\n",
      "\n",
      "\n",
      "47/47 [==============================] - 2s 49ms/step\n",
      "Optimizer: Adam\n",
      "Test loss: 0.04432033747434616\n",
      "Test accuracy: 0.9866622090339661\n",
      "Precision: 0.9996298\n",
      "Recall: 1.0\n",
      "F-score: 0.99981487\n",
      "\n",
      "\n",
      "47/47 [==============================] - 2s 49ms/step\n",
      "Optimizer: Adadelta\n",
      "Test loss: 2.0464608669281006\n",
      "Test accuracy: 0.489829957485199\n",
      "Precision: 0.9178316\n",
      "Recall: 0.99703705\n",
      "F-score: 0.9557963\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# define a list of optimizers to try\n",
    "optimizers = [tf.keras.optimizers.SGD(),tf.keras.optimizers.Adam(),tf.keras.optimizers.Adadelta()]\n",
    "\n",
    "# loop through each optimizer and train the model\n",
    "for optimizer in optimizers:\n",
    "    \n",
    "    # reset the model weights for each optimizer\n",
    "    model = keras.Sequential(\n",
    "        [\n",
    "            keras.Input(shape=(img_height, img_width, 1)),\n",
    "            layers.Conv2D(32, kernel_size=(3, 3), activation=\"relu\"),\n",
    "            layers.MaxPooling2D(pool_size=(2, 2)),\n",
    "            layers.Flatten(),\n",
    "            layers.Dense(128, activation=\"relu\"),\n",
    "            layers.Dense(train_generator.num_classes, activation=\"softmax\"),\n",
    "        ]\n",
    "    )\n",
    "    \n",
    "    # compile the model with the current optimizer\n",
    "    model.compile(optimizer=optimizer, loss=\"categorical_crossentropy\", metrics=[\"accuracy\"])\n",
    "    \n",
    "    # train the model\n",
    "    history = model.fit(train_generator, epochs=10, validation_data=test_generator, verbose=0)\n",
    "    \n",
    "    # evaluate the model on test set\n",
    "    test_loss, test_acc = model.evaluate(test_generator, verbose=0)\n",
    "    test_predictions = model.predict(test_generator)\n",
    "    test_pred_labels = np.argmax(test_predictions, axis=1)\n",
    "    test_true_labels = test_generator.classes\n",
    "    \n",
    "    # calculate precision, recall, and F-score\n",
    "    precision = tf.keras.metrics.Precision()(test_true_labels, test_pred_labels)\n",
    "    recall = tf.keras.metrics.Recall()(test_true_labels, test_pred_labels)\n",
    "    f1_score = 2 * (precision * recall) / (precision + recall)\n",
    "    \n",
    "    # print the evaluation metrics\n",
    "    print(\"Optimizer:\", optimizer.__class__.__name__)\n",
    "    print(\"Test loss:\", test_loss)\n",
    "    print(\"Test accuracy:\", test_acc)\n",
    "    print(\"Precision:\", precision.numpy())\n",
    "    print(\"Recall:\", recall.numpy())\n",
    "    print(\"F-score:\", f1_score.numpy())\n",
    "    print(\"\\n\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "41902f92",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "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": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}