Feature extraction and selection.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Num GPUs Available:  1\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "print(\"Num GPUs Available: \", len(tf.config.list_physical_devices('GPU')))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "id": "nReo9QEcE89e"
   },
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow.keras import models, layers\n",
    "import matplotlib.pyplot as plt\n",
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "from tensorflow.keras.applications.vgg16 import VGG16\n",
    "from tensorflow.keras.models import Model\n",
    "from tensorflow.keras.applications.vgg16 import preprocess_input\n",
    "from sklearn.svm import SVC\n",
    "import numpy as np\n",
    "from sklearn.model_selection import train_test_split\n",
    "from keras.models import Sequential\n",
    "from tensorflow.keras.layers import BatchNormalization\n",
    "from keras.layers import Conv2D\n",
    "from keras.layers import Input, Conv2D, MaxPooling2D, Flatten, Dense\n",
    "from keras.models import Sequential, Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "id": "pXKNNFPGFwlR"
   },
   "outputs": [],
   "source": [
    "BATCH_SIZE = 32\n",
    "IMAGE_SIZE = 224\n",
    "CHANNELS=3\n",
    "EPOCHS=10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "0gRgQ6saF1RD",
    "outputId": "9795ca02-4fa9-4ece-bb4f-c832da07db26"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 1507 images belonging to 3 classes.\n",
      "Found 645 images belonging to 3 classes.\n"
     ]
    }
   ],
   "source": [
    "data_generator = ImageDataGenerator(\n",
    "    preprocessing_function=preprocess_input,\n",
    "    validation_split=0.3) \n",
    "\n",
    "train_generator = data_generator.flow_from_directory(\n",
    "   r\"C:\\Users\\ADMIN\\Downloads\\potato\", \n",
    "    target_size =(IMAGE_SIZE, IMAGE_SIZE), \n",
    "    batch_size = BATCH_SIZE,\n",
    "    shuffle = True,\n",
    "    class_mode = 'categorical',\n",
    "    subset='training',\n",
    "    seed = 42,)\n",
    "\n",
    "validation_generator = data_generator.flow_from_directory( \n",
    "    r\"C:\\Users\\ADMIN\\Downloads\\potato\", \n",
    "    target_size =(IMAGE_SIZE, IMAGE_SIZE), \n",
    "    batch_size =BATCH_SIZE,\n",
    "    shuffle = True,\n",
    "    class_mode = 'categorical',\n",
    "    subset='validation',\n",
    "    seed = 42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "resize_and_rescale = tf.keras.Sequential([\n",
    "  layers.experimental.preprocessing.Resizing(IMAGE_SIZE, IMAGE_SIZE),\n",
    "  layers.experimental.preprocessing.Rescaling(1./255),\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "input_shape = (BATCH_SIZE, IMAGE_SIZE, IMAGE_SIZE, CHANNELS)\n",
    "n_classes = 3\n",
    "\n",
    "model = models.Sequential([\n",
    "    resize_and_rescale,\n",
    "    layers.Conv2D(32, kernel_size = (3,3), activation='relu', input_shape=input_shape),\n",
    "    layers.MaxPooling2D((2, 2)),\n",
    "    layers.Conv2D(64,  kernel_size = (3,3), activation='relu'),\n",
    "    layers.MaxPooling2D((2, 2)),\n",
    "    layers.Conv2D(64,  kernel_size = (3,3), activation='relu'),\n",
    "    layers.MaxPooling2D((2, 2)),\n",
    "    layers.Conv2D(64, (3, 3), activation='relu'),\n",
    "    layers.MaxPooling2D((2, 2)),\n",
    "    layers.Conv2D(64, (3, 3), activation='relu'),\n",
    "    layers.MaxPooling2D((2, 2)),\n",
    "    layers.Conv2D(64, (3, 3), activation='relu'),\n",
    "    layers.MaxPooling2D((2, 2)),\n",
    "    layers.Flatten(),\n",
    "    layers.Dense(64, activation='relu'),\n",
    "    layers.Dense(n_classes, activation='softmax'),\n",
    "])\n",
    "\n",
    "model.build(input_shape=input_shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(optimizer='adam',\n",
    "              loss='categorical_crossentropy',\n",
    "              metrics=['accuracy'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "48/48 [==============================] - 6s 118ms/step - loss: 0.4878 - accuracy: 0.8281 - val_loss: 0.5278 - val_accuracy: 0.7969\n",
      "Epoch 2/10\n",
      "48/48 [==============================] - 5s 109ms/step - loss: 0.4376 - accuracy: 0.8388 - val_loss: 0.4384 - val_accuracy: 0.8512\n",
      "Epoch 3/10\n",
      "48/48 [==============================] - 5s 110ms/step - loss: 0.3940 - accuracy: 0.8474 - val_loss: 0.4077 - val_accuracy: 0.8496\n",
      "Epoch 4/10\n",
      "48/48 [==============================] - 5s 111ms/step - loss: 0.3344 - accuracy: 0.8719 - val_loss: 0.3535 - val_accuracy: 0.8574\n",
      "Epoch 5/10\n",
      "48/48 [==============================] - 5s 100ms/step - loss: 0.2845 - accuracy: 0.8971 - val_loss: 0.2916 - val_accuracy: 0.8853\n",
      "Epoch 6/10\n",
      "48/48 [==============================] - 5s 103ms/step - loss: 0.3313 - accuracy: 0.8759 - val_loss: 0.3103 - val_accuracy: 0.8868\n",
      "Epoch 7/10\n",
      "48/48 [==============================] - 5s 103ms/step - loss: 0.2422 - accuracy: 0.9098 - val_loss: 0.2918 - val_accuracy: 0.8791\n",
      "Epoch 8/10\n",
      "48/48 [==============================] - 5s 107ms/step - loss: 0.1990 - accuracy: 0.9257 - val_loss: 0.3621 - val_accuracy: 0.8760\n",
      "Epoch 9/10\n",
      "48/48 [==============================] - 5s 112ms/step - loss: 0.2175 - accuracy: 0.9164 - val_loss: 0.2572 - val_accuracy: 0.8930\n",
      "Epoch 10/10\n",
      "48/48 [==============================] - 5s 106ms/step - loss: 0.1585 - accuracy: 0.9376 - val_loss: 0.2578 - val_accuracy: 0.9023\n"
     ]
    }
   ],
   "source": [
    "history=model.fit(train_generator, epochs=10, validation_data=validation_generator)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "acc = history.history['accuracy']\n",
    "val_acc = history.history['val_accuracy']\n",
    "\n",
    "loss = history.history['loss']\n",
    "val_loss = history.history['val_loss']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(8, 8))\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.plot(range(EPOCHS), acc, label='Training Accuracy')\n",
    "plt.plot(range(EPOCHS), val_acc, label='Validation Accuracy')\n",
    "plt.legend(loc='lower right')\n",
    "plt.title('Training and Validation Accuracy')\n",
    "\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.plot(range(EPOCHS), loss, label='Training Loss')\n",
    "plt.plot(range(EPOCHS), val_loss, label='Validation Loss')\n",
    "plt.legend(loc='upper right')\n",
    "plt.title('Training and Validation Loss')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " sequential_input (InputLaye  [(None, None, None, None  0        \n",
      " r)                          )]                                  \n",
      "                                                                 \n",
      " sequential (Sequential)     (None, 224, 224, None)    0         \n",
      "                                                                 \n",
      " conv2d (Conv2D)             (None, 222, 222, 32)      896       \n",
      "                                                                 \n",
      " max_pooling2d (MaxPooling2D  (None, 111, 111, 32)     0         \n",
      " )                                                               \n",
      "                                                                 \n",
      " conv2d_1 (Conv2D)           (None, 109, 109, 64)      18496     \n",
      "                                                                 \n",
      " max_pooling2d_1 (MaxPooling  (None, 54, 54, 64)       0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " conv2d_2 (Conv2D)           (None, 52, 52, 64)        36928     \n",
      "                                                                 \n",
      " max_pooling2d_2 (MaxPooling  (None, 26, 26, 64)       0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " conv2d_3 (Conv2D)           (None, 24, 24, 64)        36928     \n",
      "                                                                 \n",
      " max_pooling2d_3 (MaxPooling  (None, 12, 12, 64)       0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " conv2d_4 (Conv2D)           (None, 10, 10, 64)        36928     \n",
      "                                                                 \n",
      " max_pooling2d_4 (MaxPooling  (None, 5, 5, 64)         0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " conv2d_5 (Conv2D)           (None, 3, 3, 64)          36928     \n",
      "                                                                 \n",
      " max_pooling2d_5 (MaxPooling  (None, 1, 1, 64)         0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " flatten (Flatten)           (None, 64)                0         \n",
      "                                                                 \n",
      " dense (Dense)               (None, 64)                4160      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 171,264\n",
      "Trainable params: 171,264\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "new_model = Model(inputs=model.input, outputs=model.layers[-2].output)\n",
    "new_model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "WfpQNbo2Grjq",
    "outputId": "7777d7bc-d99a-4804-aa9d-efdcd4b36af9"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "48/48 [==============================] - 4s 81ms/step\n"
     ]
    }
   ],
   "source": [
    "train_features = model.predict(train_generator)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "0CmyNYSNPH1W",
    "outputId": "0800989e-db3f-47d6-966c-5293a915b2e2"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "21/21 [==============================] - 2s 70ms/step\n"
     ]
    }
   ],
   "source": [
    "validation_features = model.predict(validation_generator)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "8glgN6wqmN4H",
    "outputId": "e2bd10ea-3e8a-48db-af5d-05d1a067fb26"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1507, 7, 7, 512)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_features.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "h-jNGP1Dx0mF",
    "outputId": "985b6ecd-1569-4ac0-ba9c-83e3b9b2093b"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(645, 7, 7, 512)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "validation_features.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "825z8hs-uPcv",
    "outputId": "cf9cc6e4-efe7-4222-b2e9-1240c3ee30ab"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1507"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(train_generator.labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "ePZzvFIHyGMT",
    "outputId": "129cda59-f20d-4890-ef69-f4b13700bd92"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "645"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(validation_generator.labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "id": "TxLA9JdqKMy5"
   },
   "outputs": [],
   "source": [
    "x_train, x_test, y_train, y_test = train_features,validation_features,train_generator.labels,validation_generator.labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "MSaTq753581d",
    "outputId": "0dc2fcdf-d69c-4091-c111-7a6335da7916"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1507"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "VfwhZR7A6DBE",
    "outputId": "dd0a698f-e538-4613-fd53-4559675a4684"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "645"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "9ITuhyD5262z",
    "outputId": "0019fe38-1bc5-4a37-a14e-025cb11554d5"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(7, 7, 512)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train.shape[1:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "id": "2wDJKG-hPgH-"
   },
   "outputs": [],
   "source": [
    "n_samples = x_train.shape[0]\n",
    "n_features = np.prod(x_train.shape[1:])\n",
    "x_train = np.reshape(x_train, (n_samples, n_features))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "tMx8A7ce3Aix",
    "outputId": "6057bb5d-d059-4e23-a9c4-26e545224fe7"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1507, 25088)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "id": "eDvh96-ayl6j"
   },
   "outputs": [],
   "source": [
    "n_samples1 = x_test.shape[0]\n",
    "n_features1 = np.prod(x_test.shape[1:])\n",
    "x_test = np.reshape(x_test, (n_samples1, n_features1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "fH1zmYz01ozr",
    "outputId": "7a238748-4e44-4a8a-fa31-e3708cd4e115"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "51200"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_features1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_traindf = pd.DataFrame(x_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_testdf = pd.DataFrame(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.svm import SVC\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import random\n",
    "import math \n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.cluster import KMeans"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.4883720930232558\n"
     ]
    }
   ],
   "source": [
    "rf = RandomForestClassifier()\n",
    "rf.fit(x_train, y_train)\n",
    "y_val_pred = rf.predict(x_test)\n",
    "print( accuracy_score(y_test, y_val_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def fitness(features_mask):\n",
    "    # select the features based on the binary mask\n",
    "    selected_features = x_train[:, features_mask.astype(bool)]\n",
    "    # train and evaluate a KNN classifier using the selected features\n",
    "    #print(len(selected_features))\n",
    "    #knn = KNeighborsClassifier(n_neighbors=204)\n",
    "    #kmeans = KMeans(n_clusters=i)\n",
    "    #kmeans.fit(data)\n",
    "    rf = RandomForestClassifier()\n",
    "    rf.fit(selected_features, y_train)\n",
    "    #svm = SVC(kernel=\"linear\", random_state=42)\n",
    "    #svm.fit(selected_features, y_train)\n",
    "    #knn.fit(selected_features, y_train)\n",
    "    #print(\"hi\")\n",
    "    #y_val_pred = knn.predict(x_test[:, features_mask.astype(bool)])\n",
    "    #y_val_pred = svm.predict(x_test[:, features_mask.astype(bool)])\n",
    "    y_val_pred = rf.predict(x_test[:, features_mask.astype(bool)])\n",
    "    #print(len(y_val_pred))\n",
    "    #print(accuracy_score(y_test, y_val_pred))\n",
    "    return accuracy_score(y_test, y_val_pred)\n",
    "\n",
    "    \n",
    "######### Moth Flame Algo ############\n",
    "def MFO(x_train, y_train, x_test, y_test, Max_iteration,n):\n",
    "    \n",
    "    length=len(x_train.iloc[0])\n",
    "    size1=int(length)\n",
    "    \n",
    "    moth_population=np.zeros(shape=(n,size1))\n",
    "    fitness_of_moths = np.zeros(n)\n",
    "\n",
    "    ########## Initialisation of solutions randomly ##############\n",
    "    for i in range(0,n):\n",
    "            # for j in range(0,size):\n",
    "            #     if j==0:\n",
    "            #         r=random.randint(0,length-1)\n",
    "            #         while( (length-1-r)<(size-j-1) ):\n",
    "            #             r=random.randint(0,length-1)\n",
    "            #         moth_population[i][j]=r\n",
    "            #     else:\n",
    "            #         r=random.randint(0,length-1)\n",
    "            #         while( ((length-1-r)<(size-j-1) ) or moth_population[i][j-1]>=r):\n",
    "            #             r=random.randint(0,length-1)\n",
    "                    # moth_population[i][j]=r\n",
    "                    \n",
    "        moth_population[i]=np.random.randint(2, size=size1)\n",
    "    previous_population=0\n",
    "    previous_fitness=0\n",
    "    \n",
    "    # Max_iteration=m  \n",
    "    ############ Begin of Algo ##############\n",
    "    for Iteration in range(1,Max_iteration+1):\n",
    "        \n",
    "        \n",
    "        print(str(Iteration), end=\" \")\n",
    "        Flame_no = round(n-Iteration*((n-1)/Max_iteration));\n",
    "        \n",
    "        for i in range(0,n):\n",
    "            fitness_of_moths[i] = fitness(moth_population[i]) # calculation of fitness\n",
    "        \n",
    "        \n",
    "        if Iteration==1:\n",
    "             best_flame_fitness = np.sort(fitness_of_moths)[::-1] # getting best flames\n",
    "             I = np.argsort(fitness_of_moths)[::-1]\n",
    "             best_flames = moth_population[I]\n",
    "            \n",
    "        else:\n",
    "            double_population = np.concatenate((previous_population,best_flames))\n",
    "            double_fitness=np.concatenate((previous_fitness, best_flame_fitness))\n",
    "            \n",
    "            double_fitness_sorted = np.sort(double_fitness)[::-1]\n",
    "            I=np.argsort(double_fitness)[::-1]\n",
    "            double_sorted_population = double_population[I]\n",
    "            \n",
    "            best_flame_fitness = double_fitness_sorted[0:n] # getting best flames\n",
    "            best_flames = double_sorted_population[0:n]\n",
    "            \n",
    "        Best_flame_score=best_flame_fitness[0] # the best flame\n",
    "        print(str(Best_flame_score), end=\" \")\n",
    "        Best_flame_pos=best_flames[0]\n",
    "      \n",
    "        previous_population=moth_population\n",
    "        previous_fitness=fitness_of_moths\n",
    "            \n",
    "        a=-1 + Iteration*((-1)/Max_iteration);   \n",
    "        \n",
    "        for i in range(0,n):\n",
    "            for j in range(0,size1):\n",
    "                \n",
    "                if i<=Flame_no:\n",
    "                    distance_to_flame = abs(best_flames[i][j] - moth_population[i][j])\n",
    "                    b = 1\n",
    "                    t = (a-1)*random.random() + 1\n",
    "                    moth_population[i][j] = round(distance_to_flame*math.exp(b*t)*math.cos(t*2*math.pi) + best_flames[i][j]) # updating moth population\n",
    "                    \n",
    "                    if moth_population[i][j]<0: # Bringinging back to search space\n",
    "                        moth_population[i][j]=0\n",
    "                    \n",
    "                    if moth_population[i][j]>length-1: # Bringinging back to search space\n",
    "                        moth_population[i][j]=length-1\n",
    "                    \n",
    "                else:\n",
    "                    distance_to_flame = abs(best_flames[i][j] - moth_population[i][j])\n",
    "                    b = 1\n",
    "                    t = (a-1)*random.random() + 1\n",
    "                    moth_population[i][j] = round(distance_to_flame*math.exp(b*t)*math.cos(t*2*math.pi) + best_flames[Flame_no][j]) # updating moth population\n",
    "                    \n",
    "                    if moth_population[i][j]<0: # Bringinging back to search space\n",
    "                        moth_population[i][j]=0\n",
    "                        \n",
    "                    if moth_population[i][j]>length-1: # Bringinging back to search space\n",
    "                        moth_population[i][j]=length-1\n",
    "    \n",
    "        #print(\"Iteration = \" + str(Iteration))\n",
    "        #print()\n",
    "        \n",
    "    return best_flames, Best_flame_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "best_flames, Best_flame_score=MFO(x_traindf, y_train, x_testdf, y_test, 1000,5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Best_flame_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "caON2tveLnIz",
    "outputId": "e9e28987-9f5e-4fc0-8aab-f2fa956289db"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1., 1., 1., ..., 1., 1., 1.])"
      ]
     },
     "execution_count": 141,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "solution=np.ones(n_features)\n",
    "solution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 380
    },
    "id": "yZV3iQwBMOJs",
    "outputId": "86368e80-b560-4910-883d-43839c29bb4e"
   },
   "outputs": [],
   "source": [
    "# define the number of solutions in the population and the maximum number of iterations\n",
    "n_pop = 1\n",
    "max_iter = 20\n",
    "population = initialize_population(n_pop, n_features)\n",
    "best_solution,best_fitness = moth_flame_optimization(population, max_iter)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "yexHiU2rSGmh",
    "outputId": "364bd4ec-5047-47d6-f90a-c31ad6d0291d"
   },
   "outputs": [],
   "source": [
    "best_fitness"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "1xulWmz-SIVk",
    "outputId": "4da62668-58b5-4b7f-dcb1-04740de40c98"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "664"
      ]
     },
     "execution_count": 121,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "list(best_solution).count(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "7G_BF-5p2yx9"
   },
   "outputs": [],
   "source": [
    "# extract the indices of the selected features in the best solution\n",
    "selected_feature_indicestrain = [i for i in range(len(best_solution)) if best_solution[i] > 0]\n",
    "\n",
    "# extract the selected features from the original feature matrix\n",
    "selected_featurestrain = X_train[:, selected_feature_indicestrain]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "MzTlpetFmN4O"
   },
   "outputs": [],
   "source": [
    "from sklearn.metrics import accuracy_score, f1_score, precision_score, recall_score, classification_report, confusion_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "d_cE-axHmN4O",
    "outputId": "0f0b8bb8-4012-4ace-f859-09f9bfb87489"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(9529, 664)"
      ]
     },
     "execution_count": 124,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "selected_featurestrain.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "X3xtAdB-mN4P",
    "outputId": "2fabd1b0-37f3-4c28-c92d-54f06a0737c8"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 35 images belonging to 2 classes.\n"
     ]
    }
   ],
   "source": [
    "test_generator = data_generator.flow_from_directory( \n",
    "    r\"C:\\Users\\ADMIN\\Downloads\\2 test-20230502T091229Z-001\\2 test\", \n",
    "    target_size =(IMAGE_SIZE, IMAGE_SIZE), \n",
    "    batch_size =BATCH_SIZE,\n",
    "    shuffle = True,\n",
    "    class_mode = 'categorical',\n",
    "    seed = 42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "IwQ4TV64mN4Q",
    "outputId": "7b5fceb6-3985-4eca-8128-da7be23ba78f"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2/2 [==============================] - 1s 131ms/step\n"
     ]
    }
   ],
   "source": [
    "test_features = model.predict(test_generator)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "NAdBtNT6mN4Q",
    "outputId": "fbd5d0b7-e9d2-4402-c133-5f2df9b33841"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9529"
      ]
     },
     "execution_count": 135,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(train_generator.labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "qm29X8wVmN4R"
   },
   "outputs": [],
   "source": [
    "selected_feature_indices1 = [i for i in range(len(best_solution)) if best_solution[i] > 0]\n",
    "selected_features1 = test_features[:, selected_feature_indices1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "it2if9CcmN4R"
   },
   "outputs": [],
   "source": [
    "selected_features = X_train[:,selected_feature_indices1]\n",
    "knn = KNeighborsClassifier(n_neighbors=5)\n",
    "knn.fit(selected_features, y_train)\n",
    "y_pred = knn.predict(X_test[:,selected_feature_indices1])"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "provenance": []
  },
  "gpuClass": "standard",
  "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": "3.10.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}