Untitled9-4.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Untitled9.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "authorship_tag": "ABX9TyNL6i/8cvU8MvVfg91Gzk0+",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/dlguswn3659/class-MachineLearning/blob/master/Untitled9-4.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-sOoEaBxtNXA",
        "colab_type": "code",
        "outputId": "007e5427-438c-4e45-c3dd-983489607b20",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 478
        }
      },
      "source": [
        "%matplotlib inline\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import matplotlib as mpl\n",
        "import seaborn as sns\n",
        "import matplotlib.animation as animation\n",
        "import operator\n",
        "from scipy import stats \n",
        "from mpl_toolkits.mplot3d import Axes3D\n",
        "from sklearn.datasets.samples_generator import make_regression \n",
        "import csv\n",
        "import math\n",
        "# from google.colab import files\n",
        "\n",
        "# uploaded = files.upload()\n",
        "\n",
        "file_data   = \"mnist.csv\"\n",
        "handle_file = open(file_data, \"r\")\n",
        "data        = handle_file.readlines()\n",
        "handle_file.close()\n",
        "\n",
        "size_row    = 28    # height of the image\n",
        "size_col    = 28    # width of the image\n",
        "\n",
        "num_image   = len(data)\n",
        "count       = 0     # count for the number of images\n",
        "\n",
        "#\n",
        "# normalize the values of the input data to be [0, 1]\n",
        "#\n",
        "def normalize(data):\n",
        "\n",
        "    data_normalized = (data - min(data)) / (max(data) - min(data))\n",
        "\n",
        "    return(data_normalized)\n",
        "\n",
        "#\n",
        "# example of distance function between two vectors x and y\n",
        "#\n",
        "def distance(x, y):\n",
        "\n",
        "    d = (x - y) ** 2\n",
        "    s = np.sum(d)\n",
        "    # r = np.sqrt(s)\n",
        "\n",
        "    return(s)\n",
        "\n",
        "#\n",
        "# make a matrix each column of which represents an images in a vector form\n",
        "#\n",
        "list_image  = np.empty((size_row * size_col, num_image), dtype=float)\n",
        "list_label  = np.empty(num_image, dtype=int)\n",
        "\n",
        "for line in data:\n",
        "\n",
        "    line_data   = line.split(',')\n",
        "    label       = line_data[0]\n",
        "    im_vector   = np.asfarray(line_data[1:])\n",
        "    im_vector   = normalize(im_vector)\n",
        "\n",
        "    list_label[count]       = label\n",
        "    list_image[:, count]    = im_vector\n",
        "\n",
        "    count += 1\n",
        "\n",
        "#\n",
        "# plot first 150 images out of 10,000 with their labels\n",
        "#\n",
        "f1 = plt.figure(1)\n",
        "\n",
        "for i in range(150):\n",
        "\n",
        "    label       = list_label[i]\n",
        "    im_vector   = list_image[:, i]\n",
        "    im_matrix   = im_vector.reshape((size_row, size_col))\n",
        "\n",
        "    plt.subplot(10, 15, i+1)\n",
        "    plt.title(label)\n",
        "    plt.imshow(im_matrix, cmap='Greys', interpolation='None')\n",
        "\n",
        "    frame   = plt.gca()\n",
        "    frame.axes.get_xaxis().set_visible(False)\n",
        "    frame.axes.get_yaxis().set_visible(False)\n",
        "\n",
        "\n",
        "#plt.show()\n",
        "\n",
        "#\n",
        "# plot the average image of all the images for each digit\n",
        "#\n",
        "f2 = plt.figure(2)\n",
        "\n",
        "im_average  = np.zeros((size_row * size_col, 10), dtype=float)\n",
        "im_count    = np.zeros(10, dtype=int)\n",
        "\n",
        "for i in range(num_image):\n",
        "\n",
        "    im_average[:, list_label[i]] += list_image[:, i]\n",
        "    im_count[list_label[i]] += 1\n",
        "\n",
        "for i in range(10):\n",
        "\n",
        "    im_average[:, i] /= im_count[i]\n",
        "\n",
        "    plt.subplot(2, 5, i+1)\n",
        "    plt.title(i)\n",
        "    plt.imshow(im_average[:,i].reshape((size_row, size_col)), cmap='Greys', interpolation='None')\n",
        "\n",
        "    frame   = plt.gca()\n",
        "    frame.axes.get_xaxis().set_visible(False)\n",
        "    frame.axes.get_yaxis().set_visible(False)\n",
        "\n",
        "plt.show()\n",
        "\n"
      ],
      "execution_count": 197,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 150 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 10 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "b_YcDHFWIRFP",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "theta = np.random.normal(size=(size_row*size_col))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DemnhlAQP0XJ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def sigmoid(z): \n",
        "  g = 1 / (1 + np.exp(-z)) \n",
        "  \n",
        "  return g "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DRugdPmoP2G5",
        "colab_type": "code",
        "outputId": "52360a63-8430-4f62-e6a8-8984856ee95c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        }
      },
      "source": [
        "z_list = []\n",
        "for i in range(0, 10):\n",
        "  z_list.append(0.0)\n",
        "\n",
        "print(z_list)"
      ],
      "execution_count": 200,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xtbUZT3IP28F",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "for i in range(0, num_image):\n",
        "  label       = list_label[i]\n",
        "  im_vector   = list_image[:, i]\n",
        "  im_matrix   = im_vector.reshape((size_row, size_col))\n",
        "  y = 0\n",
        "  for ny in range(0, size_row):\n",
        "      for nz in range(0, size_col):\n",
        "        y = y + theta[size_col*ny + nz]*im_matrix[ny][nz]\n",
        "  # print(y)\n",
        "  z_list[label] = z_list[label] + sigmoid(y)\n",
        "  # print(sigmoid(y))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OWQYa_g6P4H2",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#평균내기\n",
        "\n",
        "# for i in range(0, 10):\n",
        "  # z_list[i] = z_list[i] / label_list[i]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WUWOjwQQTDdR",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#first hidden layer로 가기 위한 초기 fully connected u 생성\n",
        "w_tmp = np.ones(shape=(6000, 197, 785))\n",
        "v_tmp = np.ones(shape=(6000, 50, 197))\n",
        "u_tmp = np.ones(shape=(6000, 10, 50))\n",
        "\n",
        "matrix_u = np.zeros((197, 785))  #197줄\n",
        "matrix_u[0][0] = 1               #y레이어의 bias를 위한 값 1\n",
        "\n",
        "for i in range(1, 197):\n",
        "  for j in range(0, 785):\n",
        "    matrix_u[i][j] = np.random.normal()\n",
        "\n",
        "\n",
        "# matrix_u[196][0] = 0            #줄, 칸\n",
        "##############################################\n",
        "matrix_v = np.zeros((50, 197))  #50줄\n",
        "matrix_v[0][0] = 1               #y레이어의 bias를 위한 값 1\n",
        "\n",
        "for i in range(1, 50):\n",
        "  for j in range(0, 197):\n",
        "    matrix_v[i][j] = np.random.normal()\n",
        "\n",
        "\n",
        "# matrix_v[196][0] = 0            #줄, 칸\n",
        "##############################################\n",
        "matrix_w = np.zeros((10, 50))  #10줄\n",
        "# matrix_w[0][0] = 1               #y레이어의 bias를 위한 값 1\n",
        "\n",
        "for i in range(0, 10):\n",
        "  for j in range(0, 50):\n",
        "    matrix_w[i][j] = np.random.normal()\n",
        "\n",
        "w_tmp = matrix_w\n",
        "v_tmp = matrix_v\n",
        "u_tmp = matrix_u\n",
        "\n",
        "# matrix_w[196][0] = 0            #줄, 칸\n",
        "\n",
        "# print(matrix)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "idHIKSRGY0p9",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import operator\n",
        "\n",
        "y_ = np.ones((197, 1))\n",
        "\n",
        "x_list = np.ones(shape=(6000, 785, 1))\n",
        "\n",
        "y__list = np.ones(shape=(6000, 197, 1))\n",
        "z__list = np.ones(shape=(6000, 50, 1))\n",
        "h__list = np.ones(shape=(6000, 10, 1))\n",
        "\n",
        "y_list = np.ones(shape=(6000, 197, 1))\n",
        "z_list = np.ones(shape=(6000, 50, 1))\n",
        "h_list = np.ones(shape=(6000, 10, 1))\n",
        "l_list = np.zeros(shape=(6000, 10, 1))\n",
        "\n",
        "y_list_tmp = np.ones(shape=(6000, 197, 1))\n",
        "z_list_tmp = np.ones(shape=(6000, 50, 1))\n",
        "h_list_tmp = np.ones(shape=(6000, 10, 1))\n",
        "y__list_tmp = np.ones(shape=(6000, 197, 1))\n",
        "z__list_tmp = np.ones(shape=(6000, 50, 1))\n",
        "h__list_tmp = np.ones(shape=(6000, 10, 1))\n",
        "\n",
        "# w_tmp = np.ones(shape=(6000, 197, 785))\n",
        "# v_tmp = np.ones(shape=(6000, 50, 197))\n",
        "# u_tmp = np.ones(shape=(6000, 10, 50))\n",
        "\n",
        "# matrix_u = np.zeros((197, 785))  #197줄\n",
        "# matrix_u[0][0] = 1               #y레이어의 bias를 위한 값 1\n",
        "\n",
        "\n",
        "#training set 6000개에 대해서\n",
        "for i in range(0, 6000):\n",
        "  # print(i)\n",
        "  label       = list_label[i]\n",
        "  bias   = [1]       #bias 1\n",
        "  im_vector = np.concatenate((bias, list_image[:, i]), axis = None)\n",
        "\n",
        "  # print(type(list_image[: , i]))\n",
        "  # print(list_image[:, i])\n",
        "  # print(im_vector)\n",
        "  im_matrix   = im_vector.reshape((size_row*size_col + 1, 1))\n",
        "  \n",
        "  x_list[i] = im_matrix\n",
        "\n",
        "  y_ = np.dot(matrix_u, im_matrix)  #x와 가중치 u를 곱해서 y_를 만듦\n",
        "  y__list[i] = y_\n",
        "  \n",
        "  y_vector = np.ones((197,1)) #bias\n",
        "  for j in range (1, 197):\n",
        "    y_vector[j][0] = 1 / (1 + math.exp(-y_[j][0]))\n",
        "    # sigmoid(y_[j][0])  #sigmoid 함수를 적용해서 y_에서 y를 만듦\n",
        "\n",
        "  y   = y_vector.reshape((197, 1))\n",
        "  y_list[i] = y\n",
        "\n",
        "  z_  = np.dot(matrix_v, y)  #y와 가중치 v를 곱해서 z_를 만듦\n",
        "  z__list[i] = z_\n",
        "  z_vector = np.ones((50,1)) #bias\n",
        "  for j in range (1, 50):\n",
        "    z_vector[j][0] = 1 / (1 + math.exp(-z_[j][0]))  #sigmoid 함수를 적용해서 y_에서 y를 만듦\n",
        "\n",
        "  z   = z_vector.reshape((50, 1))\n",
        "  z_list[i] = z\n",
        "  \n",
        "  # print(z)\n",
        "\n",
        "  h_  = np.dot(matrix_w, z)  #z와 가중치 w를 곱해서 h_를 만듦\n",
        "  h__list[i] = h_\n",
        "  \n",
        "  h_vector = np.ones((10,1)) #bias\n",
        "  for j in range (0, 10):\n",
        "    h_vector[j][0] = 1 / (1 + math.exp(-h_[j][0]))  #sigmoid 함수를 적용해서 h_에서 h를 만듦\n",
        "\n",
        "  h   = h_vector.reshape((10, 1))\n",
        "  # tmp = h.ravel()\n",
        "  # max_value = max(tmp)\n",
        "  # print(max_value)\n",
        "  # max_index = tmp.index(max_value)\n",
        "  index, value = max(enumerate(h), key=operator.itemgetter(1))\n",
        "  h_list[i] = h\n",
        "  l_list[i][index] = 1\n",
        "  # print(l_list[i])\n",
        "  # print(h)\n",
        "\n",
        "  y_list_tmp[i] = y_list[i]\n",
        "  z_list_tmp[i] = z_list[i]\n",
        "  h_list_tmp[i] = h_list[i]\n",
        "  y__list_tmp[i] = y__list[i]\n",
        "  z__list_tmp[i] = z__list[i]\n",
        "  h__list_tmp[i] = h__list[i]\n",
        "\n",
        "# print(l_list)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "i1AbB1lMv1yY",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def objective_function(size, label_List, h_List):\n",
        "  result = 0.0\n",
        "  for i in range(0, size):\n",
        "    for k in range(0, 10):\n",
        "      result = result + ((-1)*label_List[i][k]*math.log(h_List[i][k][0]) - (1-label_List[i][k])*math.log(1-h_List[i][k][0]))\n",
        "  result = result / size\n",
        "\n",
        "  return result"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "z7uMF29V6ZAh",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def w_gredient_descent(learning_rate, k_value, n_value, size, label_List):\n",
        "  result = 0.0\n",
        "  for i in range(0, size):\n",
        "    result = result + (z_list_tmp[i][n_value][0]*(sigmoid(h_list_tmp[i][k_value][0]) - label_List[i][k_value][0]))\n",
        "  result = result / size\n",
        "  result = result * learning_rate\n",
        "  final = w_tmp[k_value][n_value] - result\n",
        "\n",
        "  return final"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FQMnniqxLxzl",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def v_gredient_descent(learning_rate, k_value, n_value, size, label_List):\n",
        "  result = 0.0\n",
        "  for i in range(0, size):\n",
        "    result = result + (y_list_tmp[i][n_value][0]*w_tmp[0][k_value]*(sigmoid(h_list_tmp[i][0][0]) - label_List[i][0][0])*(sigmoid(z__list_tmp[i][k_value][0])*(1 - sigmoid(z__list_tmp[i][k_value][0]))))\n",
        "  result = result / size\n",
        "  result = result * learning_rate\n",
        "  final = v_tmp[k_value][n_value] - result\n",
        "\n",
        "  return final"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UabSYhhIO6JU",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def u_gredient_descent(learning_rate, k_value, n_value, size, label_List):\n",
        "  result = 0.0\n",
        "  for i in range(0, size):\n",
        "    result = result + (x_list[i][n_value][0]*w_tmp[0][1]*v_tmp[1][k_value]*(sigmoid(h_list_tmp[i][0][0]) - label_List[i][0][0])*(sigmoid(z__list_tmp[i][1][0])*(1 - sigmoid(z__list_tmp[i][1][0])))*(sigmoid(y__list_tmp[i][k_value][0])*(1 - sigmoid(y__list_tmp[i][k_value][0]))))\n",
        "  result = result / size\n",
        "  result = result * learning_rate\n",
        "  final = u_tmp[k_value][n_value] - result\n",
        "\n",
        "  return final"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0vn3FkYXStur",
        "colab_type": "code",
        "outputId": "fdefa257-1a3f-4146-96e5-ef53e14fc255",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 197
        }
      },
      "source": [
        "#이제 여기에다가\n",
        "\n",
        "#iteration에 따라\n",
        "#cost function 값을 배열에 저장하고\n",
        "#세타값을 업데이트 시키고(위의 세개 함수 호출)\n",
        "#업데이트 시킨값을 다시 cost함수에 넣고\n",
        "#업데이트 시키고\n",
        "#이것을 반복하는 함수를 짠다.\n",
        "\n",
        "#아직 해야하는 것 : h값에 따라 가장 큰값의 index로 L값정하는거 위에 추가해야함.\n",
        "\n",
        "iteration = 5\n",
        "l_r = 0.1\n",
        "\n",
        "cost_history = np.zeros(iteration)\n",
        "\n",
        "total_number_of_predictions = 6000\n",
        "accuracy_history = np.zeros(iteration)\n",
        "\n",
        "# small_delta = np.ones(3)  #image_size, layer 2~4(3개)\n",
        "# greece_delta_w = np.ones(shape = (10, 50))\n",
        "# greece_delta_v = np.ones(shape = (50, 197))\n",
        "# greece_delta_u = np.ones(shape = (197, 785))\n",
        "\n",
        "label_group = np.zeros(shape = (6000, 10))\n",
        "\n",
        "for i in range(0, 6000):\n",
        "  label       = list_label[i]\n",
        "  im_vector   = list_image[:, i]\n",
        "\n",
        "  # X[i] = im_vector\n",
        "  # print(i)\n",
        "  # print(X[i])\n",
        "  label_group[i][label] = 1\n",
        "\n",
        "for t in range(0, iteration):\n",
        "  number_of_correct_predictions = 0\n",
        "  \n",
        "  print(t)\n",
        "\n",
        "  # cost_history[t] = objective_function(6000, l_list, h_list_tmp)\n",
        "  result = 0.0\n",
        "  for i in range(0, 6000):\n",
        "    for k in range(0, 10):\n",
        "      # if h_list_tmp[i][k][0] <= 0:\n",
        "      # print(h_list_tmp[i][k][0])\n",
        "      result = result + ((-1)*label_group[i][k]*math.log(h_list[i][k][0]) - (1-label_group[i][k])*math.log(1-h_list[i][k][0]))\n",
        "  result = result / 6000\n",
        "  cost_history[t] = result\n",
        "\n",
        "# (1 / (1 + math.exp(-y__list_tmp[i][m][0])))\n",
        "#############################################\n",
        "  # for m in range(0, 10):\n",
        "  #   # print(m)\n",
        "  #   for n in range(0, 50):\n",
        "  #     # matrix_w[m][n] = w_gredient_descent(l_r, m, n, 6000, l_list)\n",
        "  #     result = 0.0\n",
        "  #     for i in range(0,6000):\n",
        "  #       result = result + (z_list_tmp[i][n][0]*((1 / (1 + math.exp(-h_list_tmp[i][m][0]))) - l_list[i][m][0]))\n",
        "  #     result = result / 6000\n",
        "  #     result = result * l_r\n",
        "  #     final = w_tmp[m][n] - result\n",
        "  #     matrix_w[m][n] = final\n",
        "\n",
        "  # for m in range(0, 50):\n",
        "  #   # print(m)\n",
        "  #   for n in range(0, 197):\n",
        "  #     # matrix_v[m][n] = v_gredient_descent(l_r, m, n, 6000, l_list)\n",
        "  #     result = 0.0\n",
        "  #     for i in range(0, 6000):\n",
        "  #       result = result + (y_list_tmp[i][n][0]*w_tmp[0][m]*((1 / (1 + math.exp(-h_list_tmp[i][0][0]))) - l_list[i][0][0])*((1 / (1 + math.exp(-z__list_tmp[i][m][0])))*(1 - (1 / (1 + math.exp(-z__list_tmp[i][m][0]))))))\n",
        "  #     result = result / 6000\n",
        "  #     result = result * l_r\n",
        "  #     final = v_tmp[m][n] - result\n",
        "  #     matrix_v[m][n] = final\n",
        "\n",
        "  # for m in range(0, 197):\n",
        "  #   # print(m)\n",
        "  #   for n in range(0, 785):\n",
        "  #     # matrix_u[m][n] = u_gredient_descent(l_r, m, n, 6000, l_list)\n",
        "  #     result = 0.0\n",
        "  #     for i in range(0, 6000):\n",
        "  #       result = result + (x_list[i][n][0]*w_tmp[0][1]*v_tmp[1][m]*((1 / (1 + math.exp(-h_list_tmp[i][0][0]))) - l_list[i][0][0])*((1 / (1 + math.exp(-z__list_tmp[i][1][0])))*(1 - (1 / (1 + math.exp(-z__list_tmp[i][1][0])))))*((1 / (1 + math.exp(-y__list_tmp[i][m][0])))*(1 - (1 / (1 + math.exp(-y__list_tmp[i][m][0]))))))\n",
        "  #     result = result / 6000\n",
        "  #     result = result * l_r\n",
        "  #     final = u_tmp[m][n] - result\n",
        "  #     matrix_u[m][n] = final\n",
        "#######################################################################\n",
        "##gradient descent\n",
        "  result_w = 0\n",
        "  result_v = 0\n",
        "  result_u = 0\n",
        "\n",
        "  for i in range(0, 6000):\n",
        "    small_delta_2 = np.zeros(shape = (10, 1))\n",
        "    for j in range(0, 10):\n",
        "      small_delta_2[j][0] = h_list[i][j] - label_group[i][j]\n",
        "    # print(h_list[i])\n",
        "    # print(label_group[i])\n",
        "    # small_delta_2 = h_list[i] - label_group[i] #여기 label_group 다시 봐야함.\n",
        "    small_delta_1 = np.dot(np.transpose(matrix_w), small_delta_2)\n",
        "    small_delta_0 = np.dot(np.transpose(matrix_v), small_delta_1)*(y_list_tmp[i]*(1 - y_list_tmp[i]))\n",
        "\n",
        "    # print(small_delta_2)\n",
        "    result_w = result_w + np.dot(small_delta_2, np.transpose(z_list[i]))\n",
        "    result_v = result_v + np.dot(small_delta_1, np.transpose(y_list[i]))\n",
        "    result_u = result_u + np.dot(small_delta_0, np.transpose(x_list[i]))\n",
        "\n",
        "  matrix_w = matrix_w - (result_w / 6000 *l_r)\n",
        "  matrix_v = matrix_v - (result_v / 6000 *l_r)\n",
        "  matrix_u = matrix_u - (result_u / 6000 *l_r)\n",
        "\n",
        "\n",
        "####################################################################\n",
        "\n",
        "  for k in range(0, 6000):\n",
        "  # print(i)\n",
        "    label       = list_label[k]\n",
        "    bias   = [1]       #bias 1\n",
        "    im_vector = np.concatenate((bias, list_image[:, k]), axis = None)\n",
        "\n",
        "    # print(type(list_image[: , k]))\n",
        "    # print(list_image[:, k])\n",
        "    # print(im_vector)\n",
        "    im_matrix   = im_vector.reshape((size_row*size_col + 1, 1))\n",
        "\n",
        "    y_ = np.dot(matrix_u, im_matrix)  #x와 가중치 u를 곱해서 y_를 만듦\n",
        "    y__list[k] = y_\n",
        "    \n",
        "    y_vector = np.ones((197,1)) #bias\n",
        "    for j in range (1, 197):\n",
        "      y_vector[j][0] = 1 / (1 + math.exp(-y_[j][0]))  #sigmoid 함수를 적용해서 y_에서 y를 만듦\n",
        "    \n",
        "    # print(y_vector) #시그모이드 값 확인\n",
        "    y   = y_vector.reshape((197, 1))\n",
        "    y_list[k] = y\n",
        "\n",
        "    z_  = np.dot(matrix_v, y)  #y와 가중치 v를 곱해서 z_를 만듦\n",
        "    z__list[k] = z_\n",
        "    z_vector = np.ones((50,1)) #bias\n",
        "    for j in range (1, 50):\n",
        "      z_vector[j][0] = 1 / (1 + math.exp(-z_[j][0]))  #sigmoid 함수를 적용해서 y_에서 y를 만듦\n",
        "\n",
        "    z   = z_vector.reshape((50, 1))\n",
        "    z_list[k] = z\n",
        "    \n",
        "    # print(z)\n",
        "\n",
        "    h_  = np.dot(matrix_w, z)  #z와 가중치 w를 곱해서 h_를 만듦\n",
        "    h__list[k] = h_\n",
        "    \n",
        "    h_vector = np.ones((10,1)) #bias\n",
        "    for j in range (0, 10):\n",
        "      h_vector[j][0] = 1 / (1 + math.exp(-h_[j][0]))  #sigmoid 함수를 적용해서 h_에서 h를 만듦\n",
        "\n",
        "    h   = h_vector.reshape((10, 1))\n",
        "    # print(h)\n",
        "    # tmp = h.ravel()\n",
        "    # max_value = max(tmp)\n",
        "    # print(max_value)\n",
        "    # max_index = tmp.index(max_value)\n",
        "    index, value = max(enumerate(h), key=operator.itemgetter(1))\n",
        "    h_list[k] = h\n",
        "    l_list[k][index] = 1\n",
        "    # print(l_list[i])\n",
        "    # print(h)\n",
        "    if index == label:\n",
        "      number_of_correct_predictions = number_of_correct_predictions + 1\n",
        "\n",
        "    y_list_tmp[k] = y_list[k]\n",
        "    z_list_tmp[k] = z_list[k]\n",
        "    h_list_tmp[k] = h_list[k]\n",
        "    y__list_tmp[k] = y__list[k]\n",
        "    z__list_tmp[k] = z__list[k]\n",
        "    h__list_tmp[k] = h__list[k]\n",
        "\n",
        "\n",
        "  accuracy_history[t] = number_of_correct_predictions/total_number_of_predictions*100\n",
        "  print(accuracy_history[t])\n",
        "  t = t + 1"
      ],
      "execution_count": 209,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0\n",
            "17.066666666666666\n",
            "1\n",
            "19.21666666666667\n",
            "2\n",
            "19.133333333333333\n",
            "3\n",
            "17.866666666666667\n",
            "4\n",
            "21.633333333333333\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kFfy7dv5_YR8",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import operator\n",
        "\n",
        "y_ = np.ones((197, 1))\n",
        "\n",
        "x_list = np.ones(shape=(4000, 785, 1))\n",
        "\n",
        "y__list = np.ones(shape=(4000, 197, 1))\n",
        "z__list = np.ones(shape=(4000, 50, 1))\n",
        "h__list = np.ones(shape=(4000, 10, 1))\n",
        "\n",
        "y_list = np.ones(shape=(4000, 197, 1))\n",
        "z_list = np.ones(shape=(4000, 50, 1))\n",
        "h_list = np.ones(shape=(4000, 10, 1))\n",
        "l_list = np.zeros(shape=(4000, 10, 1))\n",
        "\n",
        "y_list_tmp = np.ones(shape=(4000, 197, 1))\n",
        "z_list_tmp = np.ones(shape=(4000, 50, 1))\n",
        "h_list_tmp = np.ones(shape=(4000, 10, 1))\n",
        "y__list_tmp = np.ones(shape=(4000, 197, 1))\n",
        "z__list_tmp = np.ones(shape=(4000, 50, 1))\n",
        "h__list_tmp = np.ones(shape=(4000, 10, 1))\n",
        "\n",
        "# w_tmp = np.ones(shape=(4000, 197, 785))\n",
        "# v_tmp = np.ones(shape=(4000, 50, 197))\n",
        "# u_tmp = np.ones(shape=(4000, 10, 50))\n",
        "\n",
        "# matrix_u = np.zeros((197, 785))  #197줄\n",
        "# matrix_u[0][0] = 1               #y레이어의 bias를 위한 값 1\n",
        "\n",
        "\n",
        "#training set 6000개에 대해서\n",
        "for i in range(0, 4000):\n",
        "  # print(i)\n",
        "  label       = list_label[i + 6000]\n",
        "  bias   = [1]       #bias 1\n",
        "  im_vector = np.concatenate((bias, list_image[:, i + 6000]), axis = None)\n",
        "\n",
        "  # print(type(list_image[: , i]))\n",
        "  # print(list_image[:, i])\n",
        "  # print(im_vector)\n",
        "  im_matrix   = im_vector.reshape((size_row*size_col + 1, 1))\n",
        "  \n",
        "  x_list[i] = im_matrix\n",
        "\n",
        "  y_ = np.dot(matrix_u, im_matrix)  #x와 가중치 u를 곱해서 y_를 만듦\n",
        "  y__list[i] = y_\n",
        "  \n",
        "  y_vector = np.ones((197,1)) #bias\n",
        "  for j in range (1, 197):\n",
        "    y_vector[j][0] = 1 / (1 + math.exp(-y_[j][0]))\n",
        "    # sigmoid(y_[j][0])  #sigmoid 함수를 적용해서 y_에서 y를 만듦\n",
        "\n",
        "  y   = y_vector.reshape((197, 1))\n",
        "  y_list[i] = y\n",
        "\n",
        "  z_  = np.dot(matrix_v, y)  #y와 가중치 v를 곱해서 z_를 만듦\n",
        "  z__list[i] = z_\n",
        "  z_vector = np.ones((50,1)) #bias\n",
        "  for j in range (1, 50):\n",
        "    z_vector[j][0] = 1 / (1 + math.exp(-z_[j][0]))  #sigmoid 함수를 적용해서 y_에서 y를 만듦\n",
        "\n",
        "  z   = z_vector.reshape((50, 1))\n",
        "  z_list[i] = z\n",
        "  \n",
        "  # print(z)\n",
        "\n",
        "  h_  = np.dot(matrix_w, z)  #z와 가중치 w를 곱해서 h_를 만듦\n",
        "  h__list[i] = h_\n",
        "  \n",
        "  h_vector = np.ones((10,1)) #bias\n",
        "  for j in range (0, 10):\n",
        "    h_vector[j][0] = 1 / (1 + math.exp(-h_[j][0]))  #sigmoid 함수를 적용해서 h_에서 h를 만듦\n",
        "\n",
        "  h   = h_vector.reshape((10, 1))\n",
        "  # tmp = h.ravel()\n",
        "  # max_value = max(tmp)\n",
        "  # print(max_value)\n",
        "  # max_index = tmp.index(max_value)\n",
        "  index, value = max(enumerate(h), key=operator.itemgetter(1))\n",
        "  h_list[i] = h\n",
        "  l_list[i][index] = 1\n",
        "  # print(l_list[i])\n",
        "  # print(h)\n",
        "\n",
        "  y_list_tmp[i] = y_list[i]\n",
        "  z_list_tmp[i] = z_list[i]\n",
        "  h_list_tmp[i] = h_list[i]\n",
        "  y__list_tmp[i] = y__list[i]\n",
        "  z__list_tmp[i] = z__list[i]\n",
        "  h__list_tmp[i] = h__list[i]\n",
        "\n",
        "# print(l_list)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nzuVCwhp-DLC",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 197
        },
        "outputId": "7fe93794-ff6a-4866-bda9-1758113f1c4b"
      },
      "source": [
        "#test data!\n",
        "\n",
        "cost_history2 = np.zeros(iteration)\n",
        "\n",
        "total_number_of_predictions2 = 4000\n",
        "accuracy_history2 = np.zeros(iteration)\n",
        "\n",
        "# small_delta = np.ones(3)  #image_size, layer 2~4(3개)\n",
        "# greece_delta_w = np.ones(shape = (10, 50))\n",
        "# greece_delta_v = np.ones(shape = (50, 197))\n",
        "# greece_delta_u = np.ones(shape = (197, 785))\n",
        "\n",
        "label_group2 = np.zeros(shape = (4000, 10))\n",
        "\n",
        "for i in range(6000, 10000):\n",
        "  label       = list_label[i]\n",
        "  im_vector   = list_image[:, i]\n",
        "\n",
        "  # X[i] = im_vector\n",
        "  # print(i)\n",
        "  # print(X[i])\n",
        "  label_group2[i - 6000][label] = 1\n",
        "\n",
        "for t in range(0, iteration):\n",
        "  number_of_correct_predictions2 = 0\n",
        "  \n",
        "  print(t)\n",
        "\n",
        "  # cost_history[t] = objective_function(6000, l_list, h_list_tmp)\n",
        "  result = 0.0\n",
        "  for i in range(0, 4000):\n",
        "    for k in range(0, 10):\n",
        "      # if h_list_tmp[i][k][0] <= 0:\n",
        "      # print(h_list_tmp[i][k][0])\n",
        "      result = result + ((-1)*label_group2[i][k]*math.log(h_list[i][k][0]) - (1-label_group2[i][k])*math.log(1-h_list[i][k][0]))\n",
        "  result = result / 4000\n",
        "  cost_history2[t] = result\n",
        "\n",
        "# (1 / (1 + math.exp(-y__list_tmp[i][m][0])))\n",
        "#############################################\n",
        "  # for m in range(0, 10):\n",
        "  #   # print(m)\n",
        "  #   for n in range(0, 50):\n",
        "  #     # matrix_w[m][n] = w_gredient_descent(l_r, m, n, 6000, l_list)\n",
        "  #     result = 0.0\n",
        "  #     for i in range(0,6000):\n",
        "  #       result = result + (z_list_tmp[i][n][0]*((1 / (1 + math.exp(-h_list_tmp[i][m][0]))) - l_list[i][m][0]))\n",
        "  #     result = result / 6000\n",
        "  #     result = result * l_r\n",
        "  #     final = w_tmp[m][n] - result\n",
        "  #     matrix_w[m][n] = final\n",
        "\n",
        "  # for m in range(0, 50):\n",
        "  #   # print(m)\n",
        "  #   for n in range(0, 197):\n",
        "  #     # matrix_v[m][n] = v_gredient_descent(l_r, m, n, 6000, l_list)\n",
        "  #     result = 0.0\n",
        "  #     for i in range(0, 6000):\n",
        "  #       result = result + (y_list_tmp[i][n][0]*w_tmp[0][m]*((1 / (1 + math.exp(-h_list_tmp[i][0][0]))) - l_list[i][0][0])*((1 / (1 + math.exp(-z__list_tmp[i][m][0])))*(1 - (1 / (1 + math.exp(-z__list_tmp[i][m][0]))))))\n",
        "  #     result = result / 6000\n",
        "  #     result = result * l_r\n",
        "  #     final = v_tmp[m][n] - result\n",
        "  #     matrix_v[m][n] = final\n",
        "\n",
        "  # for m in range(0, 197):\n",
        "  #   # print(m)\n",
        "  #   for n in range(0, 785):\n",
        "  #     # matrix_u[m][n] = u_gredient_descent(l_r, m, n, 6000, l_list)\n",
        "  #     result = 0.0\n",
        "  #     for i in range(0, 6000):\n",
        "  #       result = result + (x_list[i][n][0]*w_tmp[0][1]*v_tmp[1][m]*((1 / (1 + math.exp(-h_list_tmp[i][0][0]))) - l_list[i][0][0])*((1 / (1 + math.exp(-z__list_tmp[i][1][0])))*(1 - (1 / (1 + math.exp(-z__list_tmp[i][1][0])))))*((1 / (1 + math.exp(-y__list_tmp[i][m][0])))*(1 - (1 / (1 + math.exp(-y__list_tmp[i][m][0]))))))\n",
        "  #     result = result / 6000\n",
        "  #     result = result * l_r\n",
        "  #     final = u_tmp[m][n] - result\n",
        "  #     matrix_u[m][n] = final\n",
        "#######################################################################\n",
        "##gradient descent\n",
        "  result_w = 0\n",
        "  result_v = 0\n",
        "  result_u = 0\n",
        "\n",
        "  for i in range(0, 4000):\n",
        "    small_delta_2 = np.zeros(shape = (10, 1))\n",
        "    for j in range(0, 10):\n",
        "      small_delta_2[j][0] = h_list[i][j] - label_group[i][j]\n",
        "    # print(h_list[i])\n",
        "    # print(label_group[i])\n",
        "    # small_delta_2 = h_list[i] - label_group[i] #여기 label_group 다시 봐야함.\n",
        "    small_delta_1 = np.dot(np.transpose(matrix_w), small_delta_2)\n",
        "    small_delta_0 = np.dot(np.transpose(matrix_v), small_delta_1)*(y_list_tmp[i]*(1 - y_list_tmp[i]))\n",
        "\n",
        "    # print(small_delta_2)\n",
        "    result_w = result_w + np.dot(small_delta_2, np.transpose(z_list[i]))\n",
        "    result_v = result_v + np.dot(small_delta_1, np.transpose(y_list[i]))\n",
        "    result_u = result_u + np.dot(small_delta_0, np.transpose(x_list[i]))\n",
        "\n",
        "  matrix_w = matrix_w - (result_w / 4000 *l_r)\n",
        "  matrix_v = matrix_v - (result_v / 4000 *l_r)\n",
        "  matrix_u = matrix_u - (result_u / 4000 *l_r)\n",
        "\n",
        "\n",
        "####################################################################\n",
        "\n",
        "  for k in range(0, 4000):\n",
        "  # print(i)\n",
        "    label       = list_label[k + 6000]\n",
        "    bias   = [1]       #bias 1\n",
        "    im_vector = np.concatenate((bias, list_image[:, k + 6000]), axis = None)\n",
        "\n",
        "    # print(type(list_image[: , k]))\n",
        "    # print(list_image[:, k])\n",
        "    # print(im_vector)\n",
        "    im_matrix   = im_vector.reshape((size_row*size_col + 1, 1))\n",
        "\n",
        "    y_ = np.dot(matrix_u, im_matrix)  #x와 가중치 u를 곱해서 y_를 만듦\n",
        "    y__list[k] = y_\n",
        "    \n",
        "    y_vector = np.ones((197,1)) #bias\n",
        "    for j in range (1, 197):\n",
        "      y_vector[j][0] = 1 / (1 + math.exp(-y_[j][0]))  #sigmoid 함수를 적용해서 y_에서 y를 만듦\n",
        "    \n",
        "    # print(y_vector) #시그모이드 값 확인\n",
        "    y   = y_vector.reshape((197, 1))\n",
        "    y_list[k] = y\n",
        "\n",
        "    z_  = np.dot(matrix_v, y)  #y와 가중치 v를 곱해서 z_를 만듦\n",
        "    z__list[k] = z_\n",
        "    z_vector = np.ones((50,1)) #bias\n",
        "    for j in range (1, 50):\n",
        "      z_vector[j][0] = 1 / (1 + math.exp(-z_[j][0]))  #sigmoid 함수를 적용해서 y_에서 y를 만듦\n",
        "\n",
        "    z   = z_vector.reshape((50, 1))\n",
        "    z_list[k] = z\n",
        "    \n",
        "    # print(z)\n",
        "\n",
        "    h_  = np.dot(matrix_w, z)  #z와 가중치 w를 곱해서 h_를 만듦\n",
        "    h__list[k] = h_\n",
        "    \n",
        "    h_vector = np.ones((10,1)) #bias\n",
        "    for j in range (0, 10):\n",
        "      h_vector[j][0] = 1 / (1 + math.exp(-h_[j][0]))  #sigmoid 함수를 적용해서 h_에서 h를 만듦\n",
        "\n",
        "    h   = h_vector.reshape((10, 1))\n",
        "    # print(h)\n",
        "    # tmp = h.ravel()\n",
        "    # max_value = max(tmp)\n",
        "    # print(max_value)\n",
        "    # max_index = tmp.index(max_value)\n",
        "    index, value = max(enumerate(h), key=operator.itemgetter(1))\n",
        "    h_list[k] = h\n",
        "    l_list[k][index] = 1\n",
        "    # print(l_list[i])\n",
        "    # print(h)\n",
        "    if index == label:\n",
        "      number_of_correct_predictions2 = number_of_correct_predictions2 + 1\n",
        "\n",
        "    y_list_tmp[k] = y_list[k]\n",
        "    z_list_tmp[k] = z_list[k]\n",
        "    h_list_tmp[k] = h_list[k]\n",
        "    y__list_tmp[k] = y__list[k]\n",
        "    z__list_tmp[k] = z__list[k]\n",
        "    h__list_tmp[k] = h__list[k]\n",
        "\n",
        "\n",
        "  accuracy_history2[t] = number_of_correct_predictions2/total_number_of_predictions2*100\n",
        "  print(accuracy_history2[t])\n",
        "  t = t + 1"
      ],
      "execution_count": 211,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0\n",
            "23.3\n",
            "1\n",
            "23.225\n",
            "2\n",
            "22.5\n",
            "3\n",
            "22.0\n",
            "4\n",
            "21.275\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "81J4ESnphg--",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "wrong_images_index = np.ones(4000)\n",
        "right_images_index = np.ones(4000)\n",
        "\n",
        "wrong_images_value = np.ones(4000)\n",
        "right_images_value = np.ones(4000)\n",
        "\n",
        "wrong_images_num = 0\n",
        "right_images_num = 0\n",
        "\n",
        "for k in range(0, 4000):\n",
        "  # print(i)\n",
        "    label       = list_label[k + 6000]\n",
        "    bias   = [1]       #bias 1\n",
        "    im_vector = np.concatenate((bias, list_image[:, k + 6000]), axis = None)\n",
        "\n",
        "    # print(type(list_image[: , k]))\n",
        "    # print(list_image[:, k])\n",
        "    # print(im_vector)\n",
        "    im_matrix   = im_vector.reshape((size_row*size_col + 1, 1))\n",
        "    # print(h_list[real_iter - 1][0])\n",
        "    index, value = max(enumerate(h_list[k]), key=operator.itemgetter(1))\n",
        "    \n",
        "\n",
        "    # print(l_list[i])\n",
        "    # print(h)\n",
        "    if index == label:\n",
        "      right_images_index[right_images_num] = int(k)\n",
        "      right_images_value[right_images_num] = int(index)\n",
        "      right_images_num = right_images_num + 1\n",
        "    else:\n",
        "      wrong_images_index[wrong_images_num] = int(k)\n",
        "      wrong_images_value[wrong_images_num] = int(index)\n",
        "      wrong_images_num = wrong_images_num + 1"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pMcbWdbh98uY",
        "colab_type": "text"
      },
      "source": [
        "# **Machine Learning Assignment09** #\n",
        "\n",
        "---\n",
        "\n",
        "\n",
        "\n",
        "Name : 이현주\n",
        "\n",
        "Student # : 20184060"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "CwXihsOJgZ0R"
      },
      "source": [
        "# **1. Plot the loss curve [10pt]**\n",
        "\n",
        "- plot the training loss at every iteration of gradient descent using the training data in blue color (the first 6,000 images) [5pt]\n",
        "- plot the testing loss at every iteration of gradient descent using the testing data in red color (the rest 4,000 images) [5pt]\n",
        "- the both curves should be presented in one figure"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DTouPM3uAf8Q",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 497
        },
        "outputId": "ad6664e7-06b4-45ae-b351-81d15b49ce9f"
      },
      "source": [
        "#그 코드를 바탕으로 training loss plotting 하기.\n",
        "real_iter = iteration\n",
        "\n",
        "fig,ax = plt.subplots(figsize=(12,8))\n",
        "\n",
        "cost_tmp = np.zeros(real_iter)\n",
        "\n",
        "for i in range(0, real_iter):\n",
        "  cost_tmp[i] = cost_history[i]\n",
        "\n",
        "\n",
        "# print(cost_history[1])\n",
        "\n",
        "ax.set_ylabel('J(Theta)')\n",
        "ax.set_xlabel('Iterations')\n",
        "_=ax.plot(range(real_iter),cost_tmp,'b.', color = 'blue')\n",
        "_=ax.plot(range(real_iter),cost_history2,'b.', color = 'red')"
      ],
      "execution_count": 214,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "OiKcfzyYgaCF"
      },
      "source": [
        "# **2. Plot the accuracy curve [10pt]**\n",
        "\n",
        "- plot the training accuracy (%) at every iteration of gradient descent using the training data in blue color (the first 6,000 images) [5pt]\n",
        "- plot the testing accuracy (%) at every iteration of gradient descent using the testing data in red color (the rest 4,000 images) [5pt]\n",
        "- the both curves should be presented in one figure"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fF1EODxQAhri",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 497
        },
        "outputId": "715593ff-1943-4c7a-8c1b-365b5df159c4"
      },
      "source": [
        "#그 코드를 바탕으로 training loss plotting 하기.\n",
        "fig,ax = plt.subplots(figsize=(12,8))\n",
        "\n",
        "# print(cost_history[1])\n",
        "\n",
        "ac_tmp = np.zeros(real_iter)\n",
        "\n",
        "for i in range(0, real_iter):\n",
        "  ac_tmp[i] = accuracy_history[i]\n",
        "\n",
        "ax.set_ylabel('accuracy')\n",
        "ax.set_xlabel('Iterations')\n",
        "_=ax.plot(range(real_iter),ac_tmp,'b.', color = 'blue')\n",
        "_=ax.plot(range(real_iter),accuracy_history2,'b.', color = 'red')\n",
        "# print(ac_tmp)"
      ],
      "execution_count": 215,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "hf4SrnjSgaXp"
      },
      "source": [
        "# **3. Plot the accuracy value [4pt]**\n",
        "\n",
        "- print the final training accuracy (%) using the training data (the first 6,000 images) [2pt]\n",
        "- print the final testing accuracy (%) using the testing data (the rest 4,000 images) [2pt]"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "B2tcQNdPAluv",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 53
        },
        "outputId": "3657e6ae-cb80-48dd-bfdf-8a66590337bb"
      },
      "source": [
        "print(\"final training accuracy using the training data : \", end = '')\n",
        "print(accuracy_history[-1], end = '%\\n')\n",
        "\n",
        "print(\"final training accuracy using the testing data : \", end = '')\n",
        "print(accuracy_history2[-1], end = '%\\n')"
      ],
      "execution_count": 218,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "final training accuracy using the training data : 21.633333333333333%\n",
            "final training accuracy using the testing data : 21.275%\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "dCAltdEPhHY_"
      },
      "source": [
        "# **4. Plot the classification example [10pt]**\n",
        "\n",
        "- present 10 correctly classified testing images with their labels at the title of each sub-figure in 2x5 array [5pt]\n",
        "- present 10 misclassified testing images with their misclassified labels at the title of each sub-figure in 2x5 array [5pt]"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qxLgEta7hU9c",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 527
        },
        "outputId": "e2406b6a-dc04-4d8e-e4e2-08bdac4421c7"
      },
      "source": [
        "\n",
        "print(\"**correctly classified testing images**\")\n",
        "for i in range(10):\n",
        "    # print(int(right_images_index[i]))\n",
        "    # print(h_list[int(right_images_index[i])])\n",
        "    label       = list_label[int(right_images_index[i])]\n",
        "    im_vector   = list_image[:, int(right_images_index[i])]\n",
        "    im_matrix   = im_vector.reshape((size_row, size_col))\n",
        "\n",
        "    plt.subplot(2, 5, i+1)\n",
        "    plt.title(int(right_images_value[i]))\n",
        "    plt.imshow(im_matrix, cmap='Greys', interpolation='None')\n",
        "\n",
        "    frame   = plt.gca()\n",
        "    frame.axes.get_xaxis().set_visible(False)\n",
        "    frame.axes.get_yaxis().set_visible(False)\n",
        "\n",
        "plt.show()\n",
        "print(\"\\n\\n\\n**misclassified testing images**\")\n",
        "for i in range(10):\n",
        "    # print(int(wrong_images_index[i]))\n",
        "    # print(h_list[int(right_images_index[i])])\n",
        "    label       = list_label[int(wrong_images_index[i])]\n",
        "    im_vector   = list_image[:, int(wrong_images_index[i])]\n",
        "    im_matrix   = im_vector.reshape((size_row, size_col))\n",
        "\n",
        "    plt.subplot(2, 5, i+1)\n",
        "    plt.title(int(wrong_images_value[i]))\n",
        "    plt.imshow(im_matrix, cmap='Greys', interpolation='None')\n",
        "\n",
        "    frame   = plt.gca()\n",
        "    frame.axes.get_xaxis().set_visible(False)\n",
        "    frame.axes.get_yaxis().set_visible(False)\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": 221,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "**correctly classified testing images**\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 10 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "\n",
            "\n",
            "**misclassified testing images**\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 10 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}