Labsheet_3_pavani(19041).ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Labsheet_3 pavani(19041).ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Paavani-git/Machine_Learning/blob/main/Labsheet_3_pavani(19041).ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0DTaseERWzUJ"
      },
      "source": [
        "## Bayes Classifier\n",
        "Bayes Theorem is a method to determine conditional probabilities – that is, the probability of one event occurring given that another event has already occurred.\n",
        "\n",
        "**Prior Probability** - P($ω_j$) It is the probability of an event before new data is collected. \n",
        "**Posterior Probability** - P($x|ω_j$) It is updated probability event occuring after considering new information\n",
        "**Likelihood** - p($x|ω_j$)\n",
        "\n",
        "**Bayes Theorm:**\n",
        "$$P(ω_j|x) = \\frac{p(x|ω_j)P(\\omega_j))}{P(x)}$$\n",
        "**Bayes Classifier for multiple features and \n",
        "multi-class classification framework:**\n",
        "Let ${ω_1,ω_2,....,ω_c}$ be the set of \"C\" state of classes.\n",
        "=>d dimensional feature vector **x**.\n",
        "Steps\n",
        "1. We compute the posterior probabilty of the pattern w.r.t each of the classes \"c\".\n",
        "2. We assign the pattern to the class for which the posterior probabilty is greatest.\n",
        "\n",
        "$$ω_{test} = argmax_j P(ω|x)=\\frac{p(x|ω_j)P(\\omega_j))}{P(x)}$$\n",
        "$$P(x) = \\sum_{j=1}^{c} p(x|ω_j)P(ω_j)$$\n",
        "\n",
        "$ω_{test}$ is the class for which posterior probaility is highest.\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "iBzI-47jQHfb"
      },
      "source": [
        "**Question 1**\n",
        "\n",
        "\n",
        "\n",
        "Build a rudimentary pattern recognizer by making use of the Bayesian decision\n",
        "theory concepts discussed in class. To this goal, you are given training images of 3 characters in a folder named  **TrainCharacters.zip**. There are 200 training images of size 128 × 128 for each character class. For evaluating the classifiers, you are provided 300 test images of size 128 × 128 in a separate folder **TestCharacters.zip** .\n",
        "\n",
        "Assume the samples to be generated from a multi dimensional Gaussian distribution, having class specific mean vectors $\\boldsymbol{\\mu}_i$. Consider each of the modelling schemes for computing the covariance matrix\n",
        "\n",
        "\n",
        "\n",
        "1.   The samples of a given character class are modelled by a separate covariance matrix $\\boldsymbol{\\Sigma}_i$.\n",
        "2.   The samples across all the character classes are pooled together to generate a common non diagonal covariance matrix $\\boldsymbol{\\Sigma}$.\n",
        "\n",
        "\n",
        "\n",
        "For each scenario, build a generative Bayesian classifier using the training images and categorize the 300 character samples contained in the test folder. The mean and the covariance matrices are to be estimated from the training data\n",
        "using the Maximum Likelihood techniques. Report the individual character accuracies as well as the averaged accuracy for each of the models. Compare the results obtained for scenarios 1 and 2 and **provide inferences** for the obtained results. \n",
        "\n",
        "Employ the pixel intensity values of the 32 x 32 resized version of the original image as features (after appropriate normalization). However, note that in order to beat the curse of\n",
        "dimensionality, you have to add a regularization term of the form $\\lambda \\textbf{I}$ in the computation of the covariance matrix. Assume $\\lambda$ value to be 0.5.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DRmSKSddfJFz",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a10e84e7-db4e-4170-9973-d3b28f2d48bc"
      },
      "source": [
        "\n",
        "# Load the Drive helper and mount\n",
        "from google.colab import drive\n",
        "\n",
        "# This will prompt for authorization.\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1WoiKdP0R22h"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "import cv2 as cv\n",
        "import matplotlib.pyplot as plt\n",
        "from glob import glob\n",
        "from decimal import *\n",
        "import math\n",
        "from scipy.stats import multivariate_normal"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lywgUL4iSSgt"
      },
      "source": [
        "path1 = glob(\"/content/drive/MyDrive/TrainCharacters/1/*.jpg\")\n",
        "path2 = glob(\"/content/drive/MyDrive/TrainCharacters/2/*.jpg\")\n",
        "path3 = glob(\"/content/drive/MyDrive/TrainCharacters/3/*.jpg\")\n",
        "path_test1 = glob(\"/content/drive/MyDrive/TestCharacters/1/*.jpg\")\n",
        "path_test2 = glob(\"/content/drive/MyDrive/TestCharacters/2/*.jpg\")\n",
        "path_test3 = glob(\"/content/drive/MyDrive/TestCharacters/3/*.jpg\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IM5Hqj0dfKF6"
      },
      "source": [
        "import math\n",
        "def gauss(x,m,c):               #postierer probability gaussain varient\n",
        "  w_1 = -0.5*(math.log(np.linalg.det(c),np.e))\n",
        "  d=x.shape[0]\n",
        "  u=x-m\n",
        "  w1=np.linalg.inv(c)\n",
        "  w_1_1 = -0.5*np.matmul(u.transpose(),np.matmul(w1,u))\n",
        "  w_1_0 = -0.5*d*math.log(2*np.pi,np.e) + math.log(1/3.0,np.e)         \n",
        "  g = w_1+w_1_0+w_1_1\n",
        "  print(x)\n",
        "  print(m)\n",
        "  print(c)\n",
        "  return (g[0][0])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_-d8eVJ1uQh0"
      },
      "source": [
        "def imagepath(path,shape):\n",
        "  image = cv.imread(path)\n",
        "  image = cv.cvtColor(image, cv.COLOR_BGR2GRAY)\n",
        "  image = cv.resize(image,(32,32))\n",
        "  image = image/255.0\n",
        "  image = image.reshape((1024,1))\n",
        "  return (np.array(image))\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_M2e5OmlcNk1"
      },
      "source": [
        "def mean_(path,shape):\n",
        "  data = []\n",
        "  mean = np.zeros((shape[0]*shape[1],1))\n",
        "  for img in range(len(path)):\n",
        "     image = cv.imread(path[img])\n",
        "     image = cv.cvtColor(image, cv.COLOR_BGR2GRAY)\n",
        "     image = cv.resize(image,(32,32))\n",
        "     image = image/255.0\n",
        "     shape = image.shape\n",
        "     image = image.reshape((1024,1))\n",
        "     data.append(image)\n",
        "     mean = mean + image\n",
        "  mean = mean/200\n",
        "  return (mean,data)\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0yM5fcUJkrYa"
      },
      "source": [
        "def covar(mean,data):\n",
        "  covar = np.zeros((1024,1024))\n",
        "  data = np.array(data)\n",
        "  for i in range(200):\n",
        "    covar = covar + np.dot((data[i] - mean),(data[i]-mean).transpose())\n",
        "  covar = covar/200\n",
        "  return covar\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QE8ZK9VSen5A"
      },
      "source": [
        "shape1 = (32,32)\n",
        "mean1, data1 = mean_(path1,shape1)\n",
        "mean2, data2 = mean_(path2, shape1)\n",
        "mean3, data3 = mean_(path3, shape1)\n",
        "mean_com,data_com = mean_(path1+path2+path3,shape1)              #pooled mean\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "X_UcCZ70ec7R"
      },
      "source": [
        "reg_fact = 0.5*(np.identity(1024))\n",
        "covar1 = covar(mean1,data1)+reg_fact\n",
        "covar2 = covar(mean2,data2)+reg_fact\n",
        "covar3 = covar(mean3,data3)+reg_fact\n",
        "covar_com = covar(mean_com,data_com)+reg_fact"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "QIcMo90hqe2E",
        "outputId": "e483feb8-e0f4-45ed-c75d-84f6a4549484"
      },
      "source": [
        "count = 0\n",
        "mis_classification_1 = []\n",
        "\n",
        "for i in range(len(path_test1)):\n",
        "  sample = imagepath(path_test1[i], (32,32))\n",
        "  check = np.array([gauss(sample,mean1, covar1), gauss(sample,mean2, covar2),gauss(sample,mean3,covar3)])\n",
        "  if np.argmax(check) == 0:\n",
        "    count = count + 1\n",
        "  else:\n",
        "    mis_classification_1.append([np.argmax(check),0,path_test1[i]])\n",
        "print(count)\n",
        "count_class1=count\n",
        "\n",
        "\n",
        "for i in range(len(path_test2)):\n",
        "  sample = imagepath(path_test2[i], (32,32))\n",
        "  check = np.array([gauss(sample,mean1, covar1), gauss(sample,mean2, covar2),gauss(sample,mean3,covar3)])\n",
        "  if np.argmax(check) == 1:\n",
        "    count = count + 1\n",
        "  else:\n",
        "    mis_classification_1.append([np.argmax(check),1,path_test2[i]])\n",
        "count_class2=count-count_class1\n",
        "print(count_class2)\n",
        "\n",
        "for i in range(len(path_test3)):\n",
        "  sample = imagepath(path_test3[i], (32,32))\n",
        "  check = np.array([gauss(sample,mean1, covar1), gauss(sample,mean2, covar2),gauss(sample,mean3,covar3)])\n",
        "  if np.argmax(check) == 2:\n",
        "    count = count + 1\n",
        "  else:\n",
        "    mis_classification_1.append([np.argmax(check),2,path_test3[i]])\n",
        "count_class3=count-count_class1-count_class2\n",
        "print(count_class3)    \n",
        "print ()\n",
        "print (\"accuracy:\",(count/(len(path_test1) + len(path_test2) + len(path_test3)))*100)\n",
        "\n",
        "\n",
        "\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[0.]\n",
            " [0.]\n",
            " [0.]\n",
            " ...\n",
            " [0.]\n",
            " [0.]\n",
            " [0.]]\n",
            "[[0.]\n",
            " [0.]\n",
            " [0.]\n",
            " ...\n",
            " [0.]\n",
            " [0.]\n",
            " [0.]]\n",
            "[[0.5 0.  0.  ... 0.  0.  0. ]\n",
            " [0.  0.5 0.  ... 0.  0.  0. ]\n",
            " [0.  0.  0.5 ... 0.  0.  0. ]\n",
            " ...\n",
            " [0.  0.  0.  ... 0.5 0.  0. ]\n",
            " [0.  0.  0.  ... 0.  0.5 0. ]\n",
            " [0.  0.  0.  ... 0.  0.  0.5]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NwA8wziCCObC"
      },
      "source": [
        "## Pooled covariance"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5wIYTStpCVc9",
        "outputId": "41cab95c-002d-4b06-952f-3a39263424c1"
      },
      "source": [
        "count1 = 0\n",
        "mis_classification_2 = []\n",
        "\n",
        "for i in range(len(path_test1)):\n",
        "  sample = imagepath(path_test1[i], (32,32))\n",
        "  check = np.array([gauss(sample,mean1, covar_com), gauss(sample,mean2, covar_com),gauss(sample,mean3,covar_com)])\n",
        "  if np.argmax(check) == 0:\n",
        "    count1 = count1 + 1\n",
        "  else:\n",
        "    mis_classification_2.append([np.argmax(check),0,path_test1[i]])\n",
        "print(count1)\n",
        "\n",
        "count_class1=count1    \n",
        "for i in (range(len(path_test2))):\n",
        "  sample = imagepath(path_test2[i], (32,32))\n",
        "  check = np.array([gauss(sample,mean1, covar_com), gauss(sample,mean2, covar_com),gauss(sample,mean3,covar_com)])\n",
        "  if np.argmax(check) == 1:\n",
        "    count1= count1 + 1\n",
        "  else:\n",
        "    mis_classification_2.append([np.argmax(check),1,path_test2[i]])\n",
        "count_class2=count1-count_class1\n",
        "print(count_class2)    \n",
        "\n",
        "for i in range(len(path_test3)):\n",
        "  sample = imagepath(path_test3[i], (32,32))\n",
        "  check = np.array([gauss(sample,mean1, covar_com), gauss(sample,mean2, covar_com),gauss(sample,mean3,covar_com)])\n",
        "  if np.argmax(check) == 2:\n",
        "    count1 = count1 + 1\n",
        "  else:\n",
        "    mis_classification_2.append([np.argmax(check),2,path_test3[i]])\n",
        "\n",
        "count_class3=count1-count_class1-count_class2\n",
        "print(count_class3)     \n",
        "print ()\n",
        "print ((count1/(len(path_test1) + len(path_test2) + len(path_test3)))*100)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "89\n",
            "90\n",
            "99\n",
            "\n",
            "92.66666666666666\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Do0SN18iAF5m"
      },
      "source": [
        "import glob\n",
        "import cv2 as cv\n",
        "from glob import glob\n",
        "import numpy as np\n",
        "def imagepath(path):\n",
        "  cv_img = []\n",
        "  # c=0\n",
        "  array_1 = []\n",
        "  for img in range(len(path)):\n",
        "    n = cv.imread(path[img])\n",
        "    n = cv.cvtColor(n, cv.COLOR_BGR2GRAY)            #RGB to GRAY SCALE IMAGE CONVERSION\n",
        "    cv_img.append(n) \n",
        "    p=cv.resize(n,(32,32))\n",
        "    p = np.array(p)\n",
        "    a = np.reshape(p,(1024,1))\n",
        "    a = a/255                   #normalization\n",
        "    # np.append(array_1,a)\n",
        "    array_1.append(a)\n",
        "  return (np.array(array_1))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 625
        },
        "id": "Sp_q_sjEOpLy",
        "outputId": "6931bb43-cb1b-4175-9071-4b7505701f24"
      },
      "source": [
        "#code for Question 2 here\n",
        "\n",
        "fig = plt.figure(figsize = (10,10))\n",
        "fig.add_subplot(2,2,1)\n",
        "image = cv.imread(mis_classification_1[4][2])\n",
        "plt.imshow(image)\n",
        "plt.title(\"Predicted class: \"+str(mis_classification_1[4][0]+1)+\" | Actual class: \"+str(mis_classification_1[4][1]+1))\n",
        "fig.add_subplot(2,2,2)\n",
        "image = cv.imread(mis_classification_1[5][2])\n",
        "plt.imshow(image)\n",
        "plt.title(\"Predicted class: \"+str(mis_classification_1[5][0]+1)+\" | Actual class: \"+str(mis_classification_1[5][1]+1))\n",
        "fig.add_subplot(2,2,3)\n",
        "image = cv.imread(mis_classification_1[-5][2])\n",
        "plt.imshow(image)\n",
        "plt.title(\"Predicted class: \"+str(mis_classification_1[-5][0]+1)+\" | Actual class: \"+str(mis_classification_1[-5][1]+1))\n",
        "fig.add_subplot(2,2,4)\n",
        "image = cv.imread(mis_classification_1[-6][2])\n",
        "plt.imshow(image)\n",
        "plt.title(\"Predicted class: \"+str(mis_classification_1[-6][0]+1)+\" | Actual class: \"+str(mis_classification_1[-6][1]+1))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Predicted class: 3 | Actual class: 2')"
            ]
          },
          "metadata": {},
          "execution_count": 13
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 4 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fe0RMtAllvON"
      },
      "source": [
        "## Question 2\n",
        "\n",
        "Give 4 examples of images from the test set that are misclassified by the classifier designed. Display both the state of nature (true label) and the predicted class for each image and the possible reason behind the misclassification.\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 625
        },
        "id": "0GlyYnlOOyOY",
        "outputId": "b22f1844-4745-43a2-bc9d-40911239535b"
      },
      "source": [
        "fig = plt.figure(figsize = (10,10))\n",
        "fig.add_subplot(2,2,1)\n",
        "image = cv.imread(mis_classification_2[4][2])\n",
        "plt.imshow(image)\n",
        "plt.title(\"Predicted class: \"+str(mis_classification_2[4][0]+1)+\" | Actual class: \"+str(mis_classification_2[4][1]+1))\n",
        "fig.add_subplot(2,2,2)\n",
        "image = cv.imread(mis_classification_2[5][2])\n",
        "plt.imshow(image)\n",
        "plt.title(\"Predicted class: \"+str(mis_classification_2[5][0]+1)+\" | Actual class: \"+str(mis_classification_2[5][1]+1))\n",
        "fig.add_subplot(2,2,3)\n",
        "image = cv.imread(mis_classification_1[-5][2])\n",
        "plt.imshow(image)\n",
        "plt.title(\"Predicted class: \"+str(mis_classification_2[-5][0]+1)+\" | Actual class: \"+str(mis_classification_2[-5][1]+1))\n",
        "fig.add_subplot(2,2,4)\n",
        "image = cv.imread(mis_classification_2[-6][2])\n",
        "plt.imshow(image)\n",
        "plt.title(\"Predicted class: \"+str(mis_classification_2[-6][0]+1)+\" | Actual class: \"+str(mis_classification_2[-6][1]+1))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Predicted class: 1 | Actual class: 2')"
            ]
          },
          "metadata": {},
          "execution_count": 14
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 4 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nRE78rxVJmAM"
      },
      "source": [
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Y1p_jv3ibJTo"
      },
      "source": [
        "**Question 3**\n",
        "What is the effect of the regularization factor $\\lambda$ on the performance of the Bayes Classifier?"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "t4Ld8gofN0ph"
      },
      "source": [
        "Regularation parameter is multiplied with identity matrix and add it to covariance matrix to make it invertiable to improve classifer performance and reduce the error.If training  samples  are  limited,  we  can  add  a  regularization  parameter λI to make sure that it can be inverted."
      ]
    }
  ]
}