SVM_clsssifier.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "###**Image** **Classification** **on** **CIFAR10** **Using** **SVM** **classifier** **with** **dimensionality** **reduction** **using** **PCA**"
      ],
      "metadata": {
        "id": "ZCnOClf8ddOv"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import os\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline\n",
        "\n",
        "# Load the CIFAR10 dataset\n",
        "from tensorflow.keras.datasets import cifar10\n",
        "\n",
        "baseDir = os.path.dirname(os.path.abspath('__file__')) + '/'\n",
        "classesName = ['plane', 'car', 'bird', 'cat', 'deer', 'dog', 'frog', 'horse', 'ship', 'truck']\n",
        "\n",
        "(xTrain, yTrain), (xTest, yTest) = cifar10.load_data()\n",
        "xVal = xTrain[49000:, :].astype(float)\n",
        "yVal = np.squeeze(yTrain[49000:, :])\n",
        "xTrain = xTrain[:49000, :].astype(float)\n",
        "yTrain = np.squeeze(yTrain[:49000, :])\n",
        "yTest = np.squeeze(yTest)\n",
        "xTest = xTest.astype(float)\n",
        "\n",
        "print(xTrain.shape)\n",
        "print(yTrain.shape)\n",
        "xTrain = np.reshape(xTrain, (xTrain.shape[0], -1)) # The -1 means that the corresponding dimension is calculated from the other given dimensions.\n",
        "xVal = np.reshape(xVal, (xVal.shape[0], -1))\n",
        "xTest = np.reshape(xTest, (xTest.shape[0], -1))\n",
        "print(xTrain.shape)\n",
        "print(xTrain[0])\n",
        "\n",
        "# Limiting the dataset size for faster processing\n",
        "xTrain = xTrain[:10000, :]\n",
        "yTrain = yTrain[:10000]\n",
        "print(xTrain.shape)\n",
        "print(yTrain.shape)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lwl2YqtmIzX2",
        "outputId": "35c0572b-8d2b-4b0c-f262-918e1ff712e8"
      },
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(49000, 32, 32, 3)\n",
            "(49000,)\n",
            "(49000, 3072)\n",
            "[ 59.  62.  63. ... 123.  92.  72.]\n",
            "(10000, 3072)\n",
            "(10000,)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Cumulative Explained variance against Number Of Components\n",
        "combined=np.vstack((xTrain,xVal))\n",
        "from sklearn.decomposition import PCA\n",
        "pca = PCA().fit(combined)\n",
        "plt.plot(np.cumsum(pca.explained_variance_ratio_))\n",
        "plt.xlabel('number of components')\n",
        "plt.ylabel('cumulative explained variance');"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 449
        },
        "id": "6rH6M5nZLm-X",
        "outputId": "47f2deff-ed69-493a-a9ea-9bfb49f729fb"
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(pca.explained_variance_ratio_.cumsum()[149])\n",
        "print(pca.explained_variance_ratio_.cumsum()[499])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Z5VZN7HNMJYZ",
        "outputId": "9556fc44-7285-4b2d-8995-1ffbd2c7367f"
      },
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.9306473212588602\n",
            "0.9846410735518448\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Reducing to 150 dimensions.\n",
        "pca = PCA(n_components=150) # project from 3072 to 150 dimensions\n",
        "pca.fit(combined)\n",
        "projected = pca.transform(combined)\n",
        "print(pca.explained_variance_.shape)\n",
        "print(pca.components_.shape)\n",
        "print(combined.shape)\n",
        "print(projected.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "K6ANIFW5MMzy",
        "outputId": "0c1bce9c-757f-4ee0-fbe0-cbf952632d27"
      },
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(150,)\n",
            "(150, 3072)\n",
            "(11000, 3072)\n",
            "(11000, 150)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "x_train=projected[:10000,:]\n",
        "y_train=yTrain\n",
        "\n",
        "x_test=projected[10000:,:]\n",
        "y_test=yVal"
      ],
      "metadata": {
        "id": "Hq7xKtQhMPyq"
      },
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(x_train.shape)\n",
        "print(y_train.shape)\n",
        "print(x_test.shape)\n",
        "print(y_test.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zQZJ4b1mMaTW",
        "outputId": "9d1679dc-3a8e-4cba-e668-4e7e10d7996d"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(10000, 150)\n",
            "(10000,)\n",
            "(1000, 150)\n",
            "(1000,)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn import svm\n",
        "svc = svm.SVC(probability=False,  kernel=\"poly\", C=1)\n",
        "svc.fit(x_train, y_train)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 74
        },
        "id": "9SJCaEKaMg5o",
        "outputId": "9afd8f70-a42b-46e3-d2ba-f0887e6cf292"
      },
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "SVC(C=1, kernel='poly')"
            ],
            "text/html": [
              "<style>#sk-container-id-1 {color: black;background-color: white;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>SVC(C=1, kernel=&#x27;poly&#x27;)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" checked><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SVC</label><div class=\"sk-toggleable__content\"><pre>SVC(C=1, kernel=&#x27;poly&#x27;)</pre></div></div></div></div></div>"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Accuracy calculation for train set\n",
        "pred = svc.predict(x_train)\n",
        "acc_train = np.mean(pred == y_train)\n",
        "print('Train Accuracy = {0:f}'.format(acc_train))\n",
        "print(pred)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SswWwN9BMt5h",
        "outputId": "78a8adbe-17b1-4d82-db6b-3287e34e2b65"
      },
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Accuracy = 0.675600\n",
            "[6 9 9 ... 1 1 7]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Accuracy calculation for test set\n",
        "pred = svc.predict(x_test)\n",
        "acc_test = np.mean(pred == y_test)\n",
        "print('Test Accuracy = {0:f}'.format(acc_test))\n",
        "print(pred)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BR5_Sr7tMxpM",
        "outputId": "1b49d19b-7f1a-4e97-e57b-184cc37825c8"
      },
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Test Accuracy = 0.402000\n",
            "[8 4 2 7 2 2 0 8 2 3 7 9 9 2 4 3 2 2 8 2 2 1 4 2 8 2 9 2 2 8 4 1 2 4 3 9 2\n",
            " 7 2 3 1 6 8 8 6 1 2 4 5 2 0 2 5 4 8 4 6 9 7 2 8 2 2 9 9 9 1 0 4 3 2 7 2 4\n",
            " 4 2 7 2 3 0 6 2 1 3 1 8 8 8 2 2 8 0 2 2 6 6 1 4 4 2 8 5 8 7 2 6 4 0 8 4 4\n",
            " 7 9 3 4 1 3 2 2 6 3 4 1 9 2 9 2 2 0 3 0 2 9 2 8 6 2 9 4 7 2 4 2 2 2 6 6 1\n",
            " 9 2 1 2 0 2 0 2 2 1 9 2 8 8 7 1 0 1 8 8 0 0 3 2 2 8 0 7 9 2 2 4 8 9 9 7 8\n",
            " 2 1 4 5 2 5 8 4 8 8 8 8 3 4 7 4 8 2 7 1 1 9 2 9 6 4 9 1 2 2 2 3 9 6 8 2 2\n",
            " 3 2 2 1 4 5 2 4 1 4 1 3 1 8 0 2 8 1 2 6 4 2 1 0 8 8 2 6 2 4 2 8 1 0 2 8 9\n",
            " 5 4 6 7 2 7 2 9 5 2 1 4 3 4 9 2 2 2 6 2 9 4 2 6 2 1 4 1 9 5 6 2 8 4 4 4 9\n",
            " 8 2 2 3 7 3 4 9 5 6 8 2 8 6 2 4 4 2 2 6 4 4 4 4 6 4 8 7 4 2 4 6 2 4 8 7 2\n",
            " 5 2 4 8 1 0 2 8 4 6 4 8 2 2 0 4 2 2 2 4 4 9 7 2 2 8 4 2 9 2 7 2 1 6 7 4 9\n",
            " 9 8 8 0 8 2 2 5 2 8 3 1 6 1 2 9 3 2 8 2 7 4 3 8 6 1 4 7 2 7 8 1 4 6 1 9 4\n",
            " 3 2 1 6 1 2 9 4 8 1 4 4 3 8 6 4 0 3 4 1 8 6 8 2 2 4 6 8 1 1 5 6 0 6 0 8 8\n",
            " 6 8 8 6 6 2 2 7 3 4 7 4 6 0 4 4 7 1 4 3 6 8 0 0 3 1 3 0 7 9 2 4 2 9 2 8 6\n",
            " 4 4 1 5 2 1 4 4 1 2 1 2 5 1 3 6 0 0 6 1 8 2 4 4 1 3 7 4 6 2 2 6 4 2 4 6 0\n",
            " 4 4 9 1 3 2 3 0 9 2 2 9 8 7 8 4 9 2 2 2 2 4 2 4 8 2 6 4 9 4 6 5 6 2 4 8 9\n",
            " 2 0 3 0 0 3 2 1 4 4 4 7 8 8 2 8 7 9 2 0 2 5 6 9 5 9 6 2 1 2 1 4 4 2 6 1 8\n",
            " 7 9 1 2 9 4 4 2 4 9 6 6 1 4 9 4 2 3 6 9 2 4 2 2 1 2 3 6 2 6 4 2 6 8 3 0 8\n",
            " 7 7 8 5 2 9 3 2 6 4 6 2 2 9 2 2 6 5 3 7 2 4 4 4 6 8 2 8 8 2 4 0 2 2 8 4 2\n",
            " 6 6 3 2 3 4 9 2 4 2 8 8 2 8 3 2 9 6 0 2 2 2 4 7 8 2 6 4 9 0 8 2 6 2 2 8 1\n",
            " 2 4 9 4 6 6 4 6 8 8 2 3 5 8 7 4 4 2 4 2 4 2 2 5 2 7 2 2 5 9 1 8 0 8 7 4 0\n",
            " 2 3 2 2 8 4 8 2 1 4 2 4 8 4 2 4 6 8 4 2 6 0 3 4 8 1 1 0 1 8 3 6 7 4 4 6 6\n",
            " 1 8 3 8 2 4 9 6 1 7 2 8 5 9 7 3 7 7 1 2 3 3 7 2 7 2 6 6 8 2 2 2 7 4 2 4 2\n",
            " 9 9 0 4 4 7 1 8 5 2 2 2 4 8 7 2 8 2 4 7 2 3 6 0 8 8 2 2 8 3 8 2 7 1 2 1 7\n",
            " 8 8 2 6 2 0 8 2 4 4 8 0 0 9 1 2 5 0 2 2 1 3 2 2 1 8 8 9 9 9 2 5 0 3 5 1 8\n",
            " 2 3 4 8 4 7 0 6 0 8 4 2 2 8 4 4 9 8 2 7 9 6 8 5 5 4 6 8 2 9 2 1 2 0 3 2 0\n",
            " 2 9 3 4 6 1 9 9 3 3 2 8 3 3 4 3 2 9 7 1 3 4 9 6 4 4 4 4 3 4 8 4 4 2 2 2 4\n",
            " 6 2 5 3 6 2 8 2 3 9 8 2 2 8 4 1 6 5 8 2 6 4 2 8 4 2 2 0 2 3 2 8 2 8 4 9 8\n",
            " 2]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Reducing to 500 dimensions.\n",
        "pca = PCA(n_components=500) # project from 3072 to 500 dimensions\n",
        "pca.fit(combined)\n",
        "projected = pca.transform(combined)\n",
        "print(pca.explained_variance_.shape)\n",
        "print(pca.components_.shape)\n",
        "print(combined.shape)\n",
        "print(projected.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NagsdHf2cQrD",
        "outputId": "1408c462-5369-456b-c14e-1bf11755b9e3"
      },
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(500,)\n",
            "(500, 3072)\n",
            "(11000, 3072)\n",
            "(11000, 500)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "x_train=projected[:10000,:]\n",
        "y_train=yTrain\n",
        "\n",
        "x_test=projected[10000:,:]\n",
        "y_test=yVal"
      ],
      "metadata": {
        "id": "uxQNgslQcTKy"
      },
      "execution_count": 27,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(x_train.shape)\n",
        "print(y_train.shape)\n",
        "print(x_test.shape)\n",
        "print(y_test.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BC6zVoMmcVHO",
        "outputId": "73040f3f-6e71-444d-e350-6b3585fe904a"
      },
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(10000, 500)\n",
            "(10000,)\n",
            "(1000, 500)\n",
            "(1000,)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn import svm\n",
        "svc = svm.SVC(probability=False,  kernel=\"poly\", C=1)\n",
        "svc.fit(x_train, y_train)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 74
        },
        "id": "YRaHCP3CcZD9",
        "outputId": "0a19881e-8389-47a5-f815-58d3ec15f8ed"
      },
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "SVC(C=1, kernel='poly')"
            ],
            "text/html": [
              "<style>#sk-container-id-2 {color: black;background-color: white;}#sk-container-id-2 pre{padding: 0;}#sk-container-id-2 div.sk-toggleable {background-color: white;}#sk-container-id-2 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-2 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-2 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-2 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-2 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-2 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-2 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-2 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-2 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-2 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-2 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-2 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-2 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-2 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-2 div.sk-item {position: relative;z-index: 1;}#sk-container-id-2 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-2 div.sk-item::before, #sk-container-id-2 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-2 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-2 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-2 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-2 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-2 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-2 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-2 div.sk-label-container {text-align: center;}#sk-container-id-2 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-2 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-2\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>SVC(C=1, kernel=&#x27;poly&#x27;)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" checked><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SVC</label><div class=\"sk-toggleable__content\"><pre>SVC(C=1, kernel=&#x27;poly&#x27;)</pre></div></div></div></div></div>"
            ]
          },
          "metadata": {},
          "execution_count": 30
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Accuracy calculation for training set\n",
        "pred = svc.predict(x_train)\n",
        "acc_train = np.mean(pred == y_train)\n",
        "print('Train Accuracy = {0:f}'.format(acc_train))\n",
        "print(pred)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "v1_zJtJLccdt",
        "outputId": "668e6ffe-16f5-48f9-eb9b-e1205d6cc6dc"
      },
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Accuracy = 0.711600\n",
            "[6 9 9 ... 1 1 7]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Accuracy calculation for test set\n",
        "pred = svc.predict(x_test)\n",
        "acc_test = np.mean(pred == y_test)\n",
        "print('Test Accuracy = {0:f}'.format(acc_test))\n",
        "print(pred)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pbOPLQe0ceV7",
        "outputId": "1bc84e3b-0f05-4f11-f023-c7cf69429684"
      },
      "execution_count": 32,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Test Accuracy = 0.395000\n",
            "[8 5 2 7 2 2 0 8 2 3 4 9 9 2 4 3 2 2 8 2 2 1 4 2 8 2 9 2 2 3 4 1 4 4 3 9 2\n",
            " 7 2 3 1 6 8 8 6 1 2 7 5 2 0 4 5 4 8 4 4 9 7 2 8 2 2 9 9 9 1 0 4 3 2 7 2 4\n",
            " 4 2 7 2 3 0 6 2 1 2 1 8 8 8 2 2 8 0 2 2 6 6 1 4 4 2 8 5 8 7 2 6 4 0 8 4 4\n",
            " 7 9 3 4 1 3 2 2 6 3 4 1 9 2 9 2 2 0 3 0 2 9 2 8 6 2 4 4 7 2 4 2 2 2 6 6 1\n",
            " 9 2 1 2 0 2 0 2 2 1 1 2 8 8 7 1 0 1 8 8 8 0 3 2 2 2 0 2 9 2 2 4 8 9 9 7 8\n",
            " 2 1 4 5 2 5 8 4 8 8 8 8 3 4 4 4 8 2 7 1 1 9 2 9 6 4 9 1 2 2 2 3 9 7 8 2 2\n",
            " 3 2 2 9 4 5 2 4 1 4 1 3 1 8 0 2 8 1 2 6 4 2 1 0 8 8 2 6 2 4 2 8 1 0 2 8 9\n",
            " 5 4 6 7 2 7 2 9 3 2 1 4 3 4 9 2 2 2 6 2 9 4 2 6 2 9 4 3 9 5 6 2 8 4 4 4 9\n",
            " 8 2 2 3 7 3 2 9 5 6 8 2 8 6 2 2 4 2 2 6 4 4 4 4 6 4 8 7 4 2 4 3 2 4 8 7 2\n",
            " 5 2 4 8 1 0 2 8 2 6 4 8 2 2 0 4 2 2 2 4 2 9 7 2 2 8 4 2 9 2 7 2 1 6 4 4 9\n",
            " 9 8 8 0 8 2 2 5 2 8 3 1 6 1 2 9 3 2 8 2 7 4 3 8 6 1 4 7 2 7 8 1 4 6 1 9 4\n",
            " 2 2 1 6 1 2 9 4 8 1 4 4 3 0 6 4 0 2 4 1 8 6 8 2 2 4 6 8 1 1 5 6 0 6 0 8 8\n",
            " 6 8 8 6 6 2 2 7 3 7 7 4 6 0 4 4 7 4 4 3 6 8 0 0 3 1 3 0 7 9 2 4 2 9 2 8 6\n",
            " 6 4 1 5 2 1 4 4 1 2 1 2 5 1 3 6 0 0 6 1 8 2 4 4 1 4 7 4 6 2 2 6 4 2 4 6 0\n",
            " 4 4 9 1 3 2 3 0 9 4 2 7 8 7 8 2 9 2 2 2 2 4 2 4 8 2 6 4 8 4 6 5 6 2 4 8 9\n",
            " 2 0 3 0 0 3 2 1 4 4 4 7 8 8 2 8 7 9 2 0 2 5 6 9 5 9 6 2 9 2 1 4 4 2 6 1 8\n",
            " 7 9 1 2 9 4 4 2 4 9 6 6 1 4 7 4 2 3 6 9 2 4 2 2 1 2 3 6 2 6 2 2 6 8 3 0 8\n",
            " 7 7 8 5 2 9 3 2 7 4 6 2 2 9 9 2 6 5 2 7 2 4 4 4 6 8 2 8 8 2 4 0 2 2 8 4 2\n",
            " 6 6 3 2 9 4 9 2 4 2 8 8 2 8 3 2 9 6 0 2 2 2 4 7 8 2 6 4 9 8 8 2 6 2 2 8 1\n",
            " 2 4 9 4 6 6 3 6 8 8 2 3 5 8 7 4 4 2 4 2 4 2 2 5 2 7 2 4 5 9 6 8 0 8 7 4 0\n",
            " 2 3 5 2 8 4 8 2 1 4 2 4 8 4 2 2 6 8 4 2 6 0 3 4 8 1 1 0 1 8 2 6 7 4 4 6 6\n",
            " 2 8 3 8 2 4 9 6 1 7 2 8 5 9 7 3 7 7 1 2 3 3 7 2 7 2 6 6 8 2 2 2 7 4 2 4 2\n",
            " 9 9 0 4 4 7 1 8 5 2 2 2 4 8 7 2 8 2 4 7 2 3 6 0 8 8 2 2 8 3 8 2 7 1 2 1 7\n",
            " 8 8 2 6 2 0 8 2 4 2 8 0 0 9 1 2 5 0 2 2 2 3 2 2 1 8 8 9 9 9 2 5 0 2 5 1 8\n",
            " 2 3 4 8 4 4 0 6 0 8 4 2 2 8 4 4 9 8 2 7 9 6 8 5 5 4 6 8 2 9 2 1 2 0 3 2 0\n",
            " 2 9 3 4 6 0 9 9 7 3 2 8 3 3 4 3 2 9 7 1 3 4 9 6 4 4 4 4 3 4 8 4 4 2 2 2 4\n",
            " 6 2 5 3 6 2 8 2 3 9 8 2 2 8 4 1 3 5 8 2 6 4 4 8 4 2 2 7 3 3 2 8 2 8 4 9 8\n",
            " 2]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## **Conclusion**\n",
        "Here, we can see that the test accuracy decreases from 40.2% to 39.6% if we increase the number of components from 150 to 500.\n",
        "\n",
        "\n",
        "---\n",
        "\n"
      ],
      "metadata": {
        "id": "JgXs1pPhfMVz"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Image Retrieval"
      ],
      "metadata": {
        "id": "mDITyndtfyxL"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics.pairwise import cosine_similarity\n",
        "\n",
        "# 1. Extract features from the query image\n",
        "query_image_index = np.random.randint(0, len(xTest)) #Random image selection from test set\n",
        "query_image = xTest[query_image_index]\n",
        "query_feature = pca.transform(query_image.reshape(1, -1))  # Transform the query image using PCA\n",
        "\n",
        "# 2. Predict the class label of the query image using the trained SVM classifier\n",
        "predicted_label = svc.predict(query_feature)[0]\n",
        "\n",
        "# 3. Retrieve images from the dataset that belong to the same class as the predicted label\n",
        "relevant_indices = np.where(yTrain == predicted_label)[0]\n",
        "relevant_images = xTrain[relevant_indices]\n",
        "\n",
        "# 4. Calculate cosine similarities between the query image and relevant images\n",
        "cosine_similarities = cosine_similarity(query_feature, pca.transform(relevant_images))\n",
        "\n",
        "# 5. Sort relevant images based on cosine similarities and get the top 10\n",
        "sorted_indices = np.argsort(cosine_similarities[0])[::-1][:10]\n",
        "top_10_relevant_images = relevant_images[sorted_indices]\n",
        "\n",
        "# 6. Display the top 10 relevant images\n",
        "plt.figure(figsize=(15, 5))\n",
        "for i in range(10):\n",
        "    plt.subplot(2, 5, i + 1)\n",
        "    plt.imshow(top_10_relevant_images[i].reshape(32, 32, 3).astype(np.uint8))\n",
        "    plt.title(f'Relevant Image {i+1}')\n",
        "    plt.axis('off')\n",
        "\n",
        "# Display the query image\n",
        "plt.figure(figsize=(5, 5))\n",
        "plt.imshow(query_image.reshape(32, 32, 3).astype(np.uint8))\n",
        "plt.title('Query Image')\n",
        "plt.axis('off')\n",
        "plt.show()\n",
        "\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 855
        },
        "id": "JBvnEDbwOnhf",
        "outputId": "47e68d2c-3ce5-429c-f25d-0b82b299d42b"
      },
      "execution_count": 35,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1500x500 with 10 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 500x500 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    }
  ]
}