A1-T2-a.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from __future__ import print_function\n",
    "import keras\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Dropout, Flatten\n",
    "from keras.layers import Conv2D, MaxPooling2D\n",
    "from keras import backend as K\n",
    "import numpy as np\n",
    "from sklearn.model_selection import train_test_split\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Load Data\n",
    "images = np.load('data/images.npy')\n",
    "labels = np.load('data/labels.npy')\n",
    "\n",
    "#Shuffle\n",
    "indices = np.arange(images.shape[0])\n",
    "np.random.shuffle(indices)\n",
    "images = images[indices]\n",
    "labels = labels[indices]\n",
    "\n",
    "# input image dimensions\n",
    "img_rows, img_cols = images.shape[1], images.shape[2]\n",
    "\n",
    "if K.image_data_format() == 'channels_first':\n",
    "    images = images.reshape(images.shape[0], 1, img_rows, img_cols)\n",
    "    input_shape = (1, img_rows, img_cols)\n",
    "else:\n",
    "    images = images.reshape(images.shape[0], img_rows, img_cols, 1)\n",
    "    input_shape = (img_rows, img_cols, 1)\n",
    "\n",
    "images = images.astype('float32')\n",
    "images /= 255\n",
    "\n",
    "#80/10/10% splits for training/validation and test sets\n",
    "train_images, temp_images, train_labels, temp_labels = train_test_split(images, labels, test_size=0.2)\n",
    "val_images, test_images, val_labels, test_labels = train_test_split(temp_images, temp_labels, test_size=0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "num_classes = 24\n",
    "def re_category(labels, num_classes):\n",
    "\n",
    "    minutes = labels[0]*60+labels[1]\n",
    "    relabels = int(minutes//(12*60/num_classes))\n",
    "    return relabels\n",
    "\n",
    "train_labels2 = np.array([re_category(h, num_classes) for h in train_labels])\n",
    "val_labels2= np.array([re_category(h, num_classes) for h in val_labels])\n",
    "test_labels2 = np.array([re_category(h, num_classes) for h in test_labels])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/chenyiru/miniconda3/envs/IDL/lib/python3.10/site-packages/keras/src/layers/convolutional/base_conv.py:107: 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",
      "2024-11-13 22:14:58.407625: I metal_plugin/src/device/metal_device.cc:1154] Metal device set to: Apple M4\n",
      "2024-11-13 22:14:58.407649: I metal_plugin/src/device/metal_device.cc:296] systemMemory: 16.00 GB\n",
      "2024-11-13 22:14:58.407666: I metal_plugin/src/device/metal_device.cc:313] maxCacheSize: 5.33 GB\n",
      "2024-11-13 22:14:58.407680: I tensorflow/core/common_runtime/pluggable_device/pluggable_device_factory.cc:305] Could not identify NUMA node of platform GPU ID 0, defaulting to 0. Your kernel may not have been built with NUMA support.\n",
      "2024-11-13 22:14:58.407689: I tensorflow/core/common_runtime/pluggable_device/pluggable_device_factory.cc:271] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 0 MB memory) -> physical PluggableDevice (device: 0, name: METAL, pci bus id: <undefined>)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/30\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-11-13 22:14:59.368456: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:117] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m16s\u001b[0m 66ms/step - accuracy: 0.0429 - loss: 3.1804 - val_accuracy: 0.0422 - val_loss: 3.1781\n",
      "Epoch 2/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 65ms/step - accuracy: 0.0432 - loss: 3.1781 - val_accuracy: 0.0361 - val_loss: 3.1781\n",
      "Epoch 3/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 67ms/step - accuracy: 0.0455 - loss: 3.1778 - val_accuracy: 0.0572 - val_loss: 3.1736\n",
      "Epoch 4/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m14s\u001b[0m 64ms/step - accuracy: 0.0613 - loss: 3.1186 - val_accuracy: 0.1583 - val_loss: 2.6874\n",
      "Epoch 5/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 65ms/step - accuracy: 0.1488 - loss: 2.6683 - val_accuracy: 0.2311 - val_loss: 2.3437\n",
      "Epoch 6/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 65ms/step - accuracy: 0.2170 - loss: 2.3973 - val_accuracy: 0.2844 - val_loss: 2.1341\n",
      "Epoch 7/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 64ms/step - accuracy: 0.2746 - loss: 2.1693 - val_accuracy: 0.3506 - val_loss: 1.9006\n",
      "Epoch 8/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m14s\u001b[0m 64ms/step - accuracy: 0.3146 - loss: 1.9877 - val_accuracy: 0.4294 - val_loss: 1.7157\n",
      "Epoch 9/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m14s\u001b[0m 64ms/step - accuracy: 0.3691 - loss: 1.8000 - val_accuracy: 0.4744 - val_loss: 1.5570\n",
      "Epoch 10/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m14s\u001b[0m 64ms/step - accuracy: 0.4249 - loss: 1.6249 - val_accuracy: 0.5422 - val_loss: 1.3617\n",
      "Epoch 11/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m14s\u001b[0m 64ms/step - accuracy: 0.4568 - loss: 1.5008 - val_accuracy: 0.5772 - val_loss: 1.2256\n",
      "Epoch 12/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m14s\u001b[0m 64ms/step - accuracy: 0.5202 - loss: 1.3330 - val_accuracy: 0.6167 - val_loss: 1.1315\n",
      "Epoch 13/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m14s\u001b[0m 64ms/step - accuracy: 0.5403 - loss: 1.2532 - val_accuracy: 0.6372 - val_loss: 1.0551\n",
      "Epoch 14/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m14s\u001b[0m 64ms/step - accuracy: 0.5750 - loss: 1.1562 - val_accuracy: 0.6722 - val_loss: 0.9513\n",
      "Epoch 15/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m14s\u001b[0m 64ms/step - accuracy: 0.6123 - loss: 1.0604 - val_accuracy: 0.7022 - val_loss: 0.8984\n",
      "Epoch 16/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m14s\u001b[0m 64ms/step - accuracy: 0.6392 - loss: 0.9890 - val_accuracy: 0.7356 - val_loss: 0.8089\n",
      "Epoch 17/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 64ms/step - accuracy: 0.6565 - loss: 0.9260 - val_accuracy: 0.7478 - val_loss: 0.7521\n",
      "Epoch 18/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 65ms/step - accuracy: 0.6806 - loss: 0.8648 - val_accuracy: 0.7694 - val_loss: 0.6907\n",
      "Epoch 19/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 65ms/step - accuracy: 0.7102 - loss: 0.8039 - val_accuracy: 0.7856 - val_loss: 0.6594\n",
      "Epoch 20/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 66ms/step - accuracy: 0.7189 - loss: 0.7516 - val_accuracy: 0.7867 - val_loss: 0.6482\n",
      "Epoch 21/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 68ms/step - accuracy: 0.7436 - loss: 0.7013 - val_accuracy: 0.7972 - val_loss: 0.6132\n",
      "Epoch 22/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 67ms/step - accuracy: 0.7586 - loss: 0.6651 - val_accuracy: 0.8139 - val_loss: 0.5822\n",
      "Epoch 23/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 65ms/step - accuracy: 0.7659 - loss: 0.6296 - val_accuracy: 0.8244 - val_loss: 0.5499\n",
      "Epoch 24/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 66ms/step - accuracy: 0.7711 - loss: 0.6080 - val_accuracy: 0.8417 - val_loss: 0.5226\n",
      "Epoch 25/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 68ms/step - accuracy: 0.7825 - loss: 0.5978 - val_accuracy: 0.8400 - val_loss: 0.4902\n",
      "Epoch 26/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 66ms/step - accuracy: 0.8082 - loss: 0.5265 - val_accuracy: 0.8389 - val_loss: 0.4924\n",
      "Epoch 27/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 65ms/step - accuracy: 0.8048 - loss: 0.5337 - val_accuracy: 0.8550 - val_loss: 0.4566\n",
      "Epoch 28/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 65ms/step - accuracy: 0.8143 - loss: 0.5039 - val_accuracy: 0.8600 - val_loss: 0.4532\n",
      "Epoch 29/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 65ms/step - accuracy: 0.8293 - loss: 0.4666 - val_accuracy: 0.8678 - val_loss: 0.4302\n",
      "Epoch 30/30\n",
      "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 65ms/step - accuracy: 0.8265 - loss: 0.4679 - val_accuracy: 0.8633 - val_loss: 0.4279\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Validation loss: 0.42789334058761597\n",
      "Validation accuracy: 0.8633333444595337\n",
      "Test loss: 0.4356205463409424\n",
      "Test accuracy: 0.8761110901832581\n"
     ]
    }
   ],
   "source": [
    "batch_size = 64\n",
    "epochs = 30\n",
    "#Build Model\n",
    "model = Sequential()\n",
    "model.add(Conv2D(32, kernel_size=(3, 3),\n",
    "                 activation='relu',\n",
    "                 input_shape=input_shape))\n",
    "model.add(Conv2D(64, (3, 3), activation='relu'))\n",
    "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "model.add(Dropout(0.1))\n",
    "model.add(Conv2D(128, (3, 3), activation='relu'))\n",
    "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "model.add(Dropout(0.1))\n",
    "model.add(Conv2D(256, (3, 3), activation='relu'))\n",
    "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "model.add(Dropout(0.1))\n",
    "model.add(Flatten())\n",
    "model.add(Dense(256, activation='relu'))\n",
    "model.add(Dropout(0.1))\n",
    "model.add(Dense(128, activation='relu'))\n",
    "model.add(Dropout(0.1))\n",
    "model.add(Dense(num_classes, activation='softmax'))\n",
    "\n",
    "model.compile(loss=keras.losses.sparse_categorical_crossentropy,\n",
    "              optimizer=keras.optimizers.Adam(learning_rate=1e-4),\n",
    "              metrics=['accuracy'])\n",
    "\n",
    "history = model.fit(train_images, train_labels2,\n",
    "          batch_size=batch_size,\n",
    "          epochs=epochs,\n",
    "          verbose=1,\n",
    "          validation_data=(val_images, val_labels2))\n",
    "df = pd.DataFrame(history.history)\n",
    "df.plot(figsize=(8, 5))\n",
    "plt.grid(True)\n",
    "plt.gca().set_ylim(0, 1)\n",
    "plt.show()\n",
    "score = model.evaluate(val_images, val_labels2, verbose=0)\n",
    "print('Validation loss:', score[0])\n",
    "print('Validation accuracy:', score[1])\n",
    "score_final = model.evaluate(test_images, test_labels2, verbose=0)\n",
    "print('Test loss:', score_final[0])\n",
    "print('Test accuracy:', score_final[1])"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "IDL",
   "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": "3.10.15"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}