ANN_mnist.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyMPH/ER5q+GDFuKuDyHS/gt",
      "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/tusharbansal22/Tensorflow-Notebooks/blob/main/ANN_mnist.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oITmJ9PhjVRu",
        "outputId": "47f0a105-ad49-4a04-dab2-665e0e09de66"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "2.9.2\n"
          ]
        }
      ],
      "source": [
        "import tensorflow as tf\n",
        "print(tf.__version__)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "mnist=tf.keras.datasets.mnist\n",
        "(x_train,y_train),(x_test,y_test)=mnist.load_data()\n",
        "x_train,x_test=x_train/255,x_test/255\n",
        "print(\" x_train shape :\",x_train.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DOruoRgAjh8x",
        "outputId": "2d200e59-8638-4662-e915-e1c80229938c"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/mnist.npz\n",
            "11490434/11490434 [==============================] - 0s 0us/step\n",
            " x_train shape : (60000, 28, 28)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model=tf.keras.models.Sequential([\n",
        "    tf.keras.layers.Flatten(input_shape=(28,28)),\n",
        "    tf.keras.layers.Dense(128,activation='relu'),\n",
        "    tf.keras.layers.Dropout(0.2),\n",
        "    tf.keras.layers.Dense(10,activation='softmax')\n",
        "])"
      ],
      "metadata": {
        "id": "VYxjaT9Lj-om"
      },
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model.compile(\n",
        "    optimizer='adam',\n",
        "    loss='sparse_categorical_crossentropy',\n",
        "    metrics=['accuracy']\n",
        ")"
      ],
      "metadata": {
        "id": "N8MDN9lVkhAI"
      },
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "r=model.fit(x_train,y_train,validation_data=(x_test,y_test),epochs=10)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qpviWH8Sk2UE",
        "outputId": "97c1118f-25bb-4aac-8539-901b6425e6aa"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/10\n",
            "1875/1875 [==============================] - 6s 3ms/step - loss: 0.0410 - accuracy: 0.9862 - val_loss: 0.0727 - val_accuracy: 0.9803\n",
            "Epoch 2/10\n",
            "1875/1875 [==============================] - 6s 3ms/step - loss: 0.0379 - accuracy: 0.9872 - val_loss: 0.0690 - val_accuracy: 0.9814\n",
            "Epoch 3/10\n",
            "1875/1875 [==============================] - 6s 3ms/step - loss: 0.0346 - accuracy: 0.9884 - val_loss: 0.0702 - val_accuracy: 0.9810\n",
            "Epoch 4/10\n",
            "1875/1875 [==============================] - 7s 4ms/step - loss: 0.0336 - accuracy: 0.9888 - val_loss: 0.0740 - val_accuracy: 0.9797\n",
            "Epoch 5/10\n",
            "1875/1875 [==============================] - 7s 4ms/step - loss: 0.0321 - accuracy: 0.9891 - val_loss: 0.0841 - val_accuracy: 0.9791\n",
            "Epoch 6/10\n",
            "1875/1875 [==============================] - 7s 4ms/step - loss: 0.0326 - accuracy: 0.9891 - val_loss: 0.0884 - val_accuracy: 0.9775\n",
            "Epoch 7/10\n",
            "1875/1875 [==============================] - 6s 3ms/step - loss: 0.0300 - accuracy: 0.9897 - val_loss: 0.0802 - val_accuracy: 0.9790\n",
            "Epoch 8/10\n",
            "1875/1875 [==============================] - 6s 3ms/step - loss: 0.0284 - accuracy: 0.9901 - val_loss: 0.0840 - val_accuracy: 0.9798\n",
            "Epoch 9/10\n",
            "1875/1875 [==============================] - 6s 3ms/step - loss: 0.0275 - accuracy: 0.9902 - val_loss: 0.0880 - val_accuracy: 0.9783\n",
            "Epoch 10/10\n",
            "1875/1875 [==============================] - 7s 4ms/step - loss: 0.0256 - accuracy: 0.9910 - val_loss: 0.0901 - val_accuracy: 0.9784\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "plt.plot(r.history['loss'],label='loss')\n",
        "plt.plot(r.history['val_loss'],label='val_loss')\n",
        "plt.legend()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "id": "vY_RjkIHlHVV",
        "outputId": "d32a60c2-b91d-4546-c8d2-b636abde0bc1"
      },
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7fd36562f820>"
            ]
          },
          "metadata": {},
          "execution_count": 9
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(r.history['accuracy'],label='acc')\n",
        "plt.plot(r.history['val_accuracy'],label='val_acc')\n",
        "plt.legend()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "id": "rbzGDTwPl88T",
        "outputId": "2135d6d6-3b2c-4df1-a1c9-ec739b8db500"
      },
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7fd36561aa60>"
            ]
          },
          "metadata": {},
          "execution_count": 10
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(model.evaluate(x_test,y_test))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-BgrditRmpbb",
        "outputId": "02acadbd-dc36-41a7-8ae3-30e039e59bfc"
      },
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "313/313 [==============================] - 1s 2ms/step - loss: 0.0901 - accuracy: 0.9784\n",
            "[0.09011079370975494, 0.9783999919891357]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import confusion_matrix\n",
        "import numpy as np\n",
        "import itertools\n",
        "\n",
        "def plot_confusion_matrix(cm,classes,normalize=False,title='Confusion Matrix',cmap=plt.cm.Blues):\n",
        "  if normalize:\n",
        "    cm=cm.astype('float')/cm.sum(axis=1)[:,np.newaxis]\n",
        "    print(\"Normalized confusion matrix\")\n",
        "  else:\n",
        "    print('Confusion matrix, without normalization')\n",
        "  print(cm)\n",
        "  plt.imshow(cm,interpolation='nearest',cmap=cmap)\n",
        "  plt.title(title)\n",
        "  plt.colorbar()\n",
        "  tick_marks=np.arange(len(classes))\n",
        "  plt.xticks(tick_marks,classes,rotation=45)\n",
        "  plt.yticks(tick_marks,classes)\n",
        "  fmt= '.2f' if normalize else 'd'\n",
        "  thresh= cm.max()/2.\n",
        "  for i,j in itertools.product(range(cm.shape[0]),range(cm.shape[1])):\n",
        "    plt.text(j,i,format(cm[i,j],fmt),\n",
        "             horizontalalignment='center',\n",
        "             color='white'if cm[i,j]>thresh else 'black')\n",
        "  plt.tight_layout()\n",
        "  plt.ylabel('True label')\n",
        "  plt.xlabel('Predicted label')\n",
        "  plt.show()\n",
        "\n",
        "p_test=model.predict(x_test).argmax(axis=1)\n",
        "cm=confusion_matrix(y_test,p_test)\n",
        "plot_confusion_matrix(cm,list(range(10)))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 525
        },
        "id": "4FvDFWl2m2r1",
        "outputId": "029760dd-4d6c-4e52-9008-92ed7adb0f66"
      },
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "313/313 [==============================] - 1s 2ms/step\n",
            "Confusion matrix, without normalization\n",
            "[[ 971    0    1    2    1    0    0    1    2    2]\n",
            " [   0 1126    3    1    0    0    1    1    3    0]\n",
            " [   3    1 1012    5    0    0    2    6    3    0]\n",
            " [   0    0    2  992    0    2    0    6    1    7]\n",
            " [   1    0    4    0  957    0    3    2    2   13]\n",
            " [   3    1    0   13    1  864    4    0    2    4]\n",
            " [   3    3    2    1    4   10  931    0    4    0]\n",
            " [   1    3   10    3    0    1    0 1004    1    5]\n",
            " [   1    0    5   10    5    3    1    4  942    3]\n",
            " [   0    2    0    4    9    2    0    7    0  985]]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAVEAAAEmCAYAAADbUaM7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOydd5gUVdaH38OQc44DEkRyziJKlCAIKqAIghjQT9w1p13jGnddVFzTuiYEc0YkqCAKKDmogCASJErOIMxwvj+qBgec0KHudPdwXp566K66de7p29Vnbt269/xEVTEMwzAiI0+sHTAMw0hkLIgahmFEgQVRwzCMKLAgahiGEQUWRA3DMKLAgqhhGEYUWBA9xRCRQiLyqYjsEZH3orAzWEQ+D9K3WCAik0RkWKz9MBIXC6JxiohcKiLzRWS/iGz2f+xnBWC6P1ABKKOqAyI1oqpvqOq5AfhzAiLSUURURD46aX8Tf//0EO3cLyLjsiunqj1VdUyE7hqGBdF4RERuBp4CHsELeNWA54C+AZg/DVipqikB2HLFNqCdiJRJt28YsDKoCsTDrn8jelTVtjjagBLAfmBAFmUK4AXZTf72FFDAP9YR2ADcAmwFNgPD/WMPAEeAo34dVwL3A+PS2a4OKJDXf385sBrYB6wBBqfbPzPdeWcC84A9/v9npjs2HXgQmOXb+Rwom8lnS/P/BWCkvy8J2AjcC0xPV3Y0sB7YCywAOvj7e5z0OZek8+Nh349DwOn+vqv8488DH6Sz/09gKiCxvi5si9/N/hLHH+2AgsBHWZT5O9AWaAo0AVoDd6c7XhEvGFfBC5TPikgpVb0Pr3f7jqoWVdWXs3JERIoATwM9VbUYXqBcnEG50sBnftkywBPAZyf1JC8FhgPlgfzArVnVDbwODPVfdwd+xPuDkZ55eG1QGngTeE9ECqrq5JM+Z5N051wGjACKAetOsncL0EhELheRDnhtN0xVbW20kSkWROOPMsB2zfp2ezDwD1Xdqqrb8HqYl6U7ftQ/flRVJ+L1xupE6M8xoKGIFFLVzaq6NIMy5wE/q+pYVU1R1beAn4A+6cq8qqorVfUQ8C5e8MsUVf0WKC0idfCC6esZlBmnqjv8Okfh9dCz+5yvqepS/5yjJ9k7iNeOTwDjgL+o6oZs7BmnOBZE448dQFkRyZtFmcqc2Ita5+87buOkIHwQKBquI6p6ALgYuBbYLCKfiUjdEPxJ86lKuvdbIvBnLHA90IkMeuYicquILPdnGuzG632Xzcbm+qwOquocvOELwQv2hpElFkTjj++A34F+WZTZhPeAKI1q/PlWN1QOAIXTva+Y/qCqTlHVbkAlvN7l/0LwJ82njRH6lMZY4Dpgot9LPI5/u307MBAopaol8cZjJc31TGxmeWsuIiPxerSbfPuGkSUWROMMVd2D9wDlWRHpJyKFRSSfiPQUkX/5xd4C7haRciJS1i+f7XSeTFgMnC0i1USkBHBX2gERqSAiff2x0d/xhgWOZWBjInCGPy0rr4hcDNQHJkToEwCqugY4B28M+GSKASl4T/Lzisi9QPF0x38DqofzBF5EzgAeAobg3dbfLiJZDjsYhgXROMQf37sZ72HRNrxb0OuBj/0iDwHzge+BH4CF/r5I6voCeMe3tYATA18e349NwE68gPZ/GdjYAfTGezCzA68H11tVt0fi00m2Z6pqRr3sKcBkvGlP64DDnHirnraQYIeILMyuHn/4ZBzwT1Vdoqo/A38DxopIgWg+g5G7EXvwaBiGETnWEzUMw4gCC6KGYRhRYEHUMAwjCiyIGoZhREFWE7pzHClQTPMUKZN9wQhoWiO7OdiGYWTGunVr2b59u2RfMnSSip+mmnIo5PJ6aNsUVe0RpA9BEFdBNE+RMhTscp8T27PeGu7ErmGcCrRv0zJwm5pyiAJ1BoZc/vDiZ+OyJxRXQdQwjFMJgVyQjdCCqGEYsUEACXSEICZYEDUMI3bkgp5o3H6C63rVZ96ofsx7oh8je9UHYMxNHfnu8fP57vHzWfZsf757/HwAShctwMT7evDb2CGMurJtVPV+PmUyjRvUoUHd03n8X49F/TnSuOaqK6hWuTwtmjYMzGYa69evp3vXTjRrXJ/mTRrwzNOjA7Xv0ndw1+YubSfq9+n6WgkPgTxJoW9xSlwG0fpVSzK8yxmcfdentL31E3q2qErNisUY9uR02t02nna3jeeTOev4ZI6Xfe3w0VQefGchf3t9XlT1pqamcuNfR/LJp5NY9P0y3nv7LZYvWxbER+KyYZfzyYTJgdg6mbx58/LYv0ax6PtlfD1zNv994dnA/Aa3vrtsc/s+c9Z2RIiEvsUpcRlE61QpybxV2zh0JJXUY8qMZVvo2/rETGsXtqvBezPXAHDw9xS++2krvx9NjareeXPnUqvW6dSoWZP8+fMz4OJLmPDpJ1HZTOOsDmdTunTpQGydTKVKlWjWvDkAxYoVo27demzaFG0Wuj9w6bvLNrfvM2dth43g3c6HusUpcenZsvW7OLNuBUoXLUCh/El0b55MlbJFjh9vX68CW/cc4pctewOtd9OmjSQnVz3+vkqVZDZujNEFFiHr1q5l8eJFtGrdJtauhITLNrfvM3a2QyOMXmgc90SdPlgSkR54YmJJwEuqGtKg1IqNe3jikx8Yf8+5HDicwvdrd3Ls2B/ZpgacVZP3Zq5243QCs3//fgYNvIjHRz1F8eLFsz/BiGtcfp9xc63EcQ8zVJwFURFJAp4FuuGpN84TkfGqGtIAzOvTfub1aT8DcP+g5mzc4SU2T8oj9G19Gu3vGB+4z5UrV2HDhj9SUm7cuIEqVapkcUb8cPToUQYNvIiLBw2m3wUXxtqdkHHZ5vZ95rztsInjHmaouPwz0BpYpaqrVfUI8DZh6KaXK14QgOSyRTi/zWm86/c8OzeuzIpNe9i082BWp0dEy1atWLXqZ9auWcORI0d47523Oa/3+YHXEzSqyrVXX0mduvW44aabY+1OWLhsc/s+c9Z2+IiNiWZDFU7MNL6BE4XLABCRESIyX0Tm6+/7j+9/49ZOzH/yAt6/oys3vzSbPQePANC/fY0Mb+WXPdufR4e1ZkjH01n5wkDqJpcI2+G8efPy5Ohn6HNed5o2qsdFAwZSv0GDsO1kxNAhg+jYoR0rV6ygVvVkXnslS7XisPh21izefGMsX381jTYtmtKmRVMmT5oYmH2Xvrtsc/s+c9Z22KRNtk/wMVFnme1FpD/QQ1Wv8t9fBrRR1eszOyepdHV1tXZ+h62dN4yIad+mJQsWzA80kuUpVlkLNBsRcvnDMx5YoKrBL+KPEpcPljYCVdO9TyZ69UfDMHINAknxO4k+VFzezs8DaotIDRHJD1wCBP80yDCMxCSXzBN11hNV1RQRuR5PlTEJeEVVl7qqzzCMBCSOxzpDxek8UVWdiKdJbhiGcRKWCs8wDCM6rCdqGIYRBdYTNQzDiJA4n/8ZKhZEDcOIHdYTNQzDiALriQZL0xplnalylmqV6UKpqNk17xlntg0j9xLs03kReQXoDWxV1Yb+vtLAO0B1YC0wUFV3iYjgZZjrBRwELlfVhf45w4C7fbMPqeqYrOpN/L60YRiJiRC0PMhrwMm69HcCU1W1NjDVfw/QE6jtbyOA5+F40L0PaIOXROk+ESmVVaUWRA3DiBHBZnFS1W+AnSft7guk9STHAP3S7X9dPWYDJUWkEtAd+EJVd6rqLuAL/hyYTyCubucNwzjFCG9MtKyIzE/3/kVVfTGbcyqo6mb/9Raggv86syxzIWWfS48FUcMwYkd4Y6Lbo8nipKoqIoGnrUvI2/lIZHBfuG8w66Y+yvz3/nZ834Vdm7Hg/b9zYMHTNK9f7fj+zm3qMuuN25n37t+Y9cbtnNPqjOPH8uVN4pm7B/H9x/ey+MO76delaUj1Hz58mLPataZ18yY0b9KABx8INuWfSwlfl5LGLu27bBOXtl1fK66/z7Bwn0/0N/82Hf//rf7+zLLMhZ19LuGCaKQyuGM/nU3fkc+esG/pL5u45Jb/MXPhLyfs37F7P/1v/C+tBj7C1feO5ZWHhh4/dsdV3dm2cx+N+/2DZhc9zIwFP4fkd4ECBZj8xTTmLlzCnPmL+XzKZObMnh3SuaHgSsLXpeywa/suZY1d2nZ5rbj+PsNCciSz/XhgmP96GPBJuv1DxaMtsMe/7Z8CnCsipfwHSuf6+zIl4YJopDK4sxb+ws49J0qKrFjzGz+v2/qnsktWbGDztj0ALPtlMwUL5CN/Pm/kY1jfdjz+yueAJ7WwY/eBkPwWEYoWLQp4GjcpR48iAc6RcyXh61J22LV9l7LGLm27vFZcf59hE2BPVETeAr4D6ojIBhG5EngM6CYiPwNd/ffgJUZaDawC/gdcB6CqO4EH8VJ5zgP+4e/LlIQbE81IBnfu3DnO6ruga1MW/7SeI0dTKFG0EAD3jexNhxa1WbNhGzc99l7ItlJTUzmzdQt++WUV1/zfSFq3iX9ZY9ftndPfZ6Lg6lqJt/YOsiOhqoMyOdQlg7IKjMzEzivAK6HW66wnKiKviMhWEfnRVR2uqVezIg/9tS/XP/Q2AHnz5iG5YilmL1nNmZf+kznfr+XRmy4I2V5SUhJzFixm1doNzJ83l6U/JmzTGI45Fa4VT2JJQt7iFZe386+RzfyqSMgpGdwq5UvyzhMjuOqesazZsB2AHbsPcODQ73w8dQkAH36xkKb1qmZlJkNKlizJOR078fnnbsbUgsR1eyeyrHFOEPS1ElftLYLkCX2LV5wF0UwmvkZNTsjglihaiA//cy33PP0J3y05UVl04jc/cnbL2gB0bF2Hn1ZvzsjEn9i2bRu7d+8G4NChQ0z98gvq1KkbqN8ucN3eiSpr7BKX10q8tbf1RAMgvWTytu3bsi0fqQzumEcvZ/qYWzjjtAqsmvwgw/q14/xOjVk1+UHaNK7Oh09fy/hnvSGSay85m1pVy3HXiJ7MfvtOZr99J+VKeQP9d4/+mLuv7cXcd+7i0vNac+cTH4X0Obds3kyPrp1o1awxZ7VrRZeu3eh1Xu+Qzg0FVxK+LmWHXdt3KWvs0rbLa8X19xkuuSGIOpNMBhCR6sCEtGQA2dGiRUudNWd+9gUjwBKQGEbkuJBMTipdQ4t2/0fI5fe+PfSUk0w2DMPIHPG3BMeCqGEYMUGI79v0UHE5xSmjia+GYRjHyQ1joi515zOb+GoYhgEEO9k+VtjtvGEYMcOCqGEYRqTYgyXDMIzIEYQ8eWI+VT1qLIgahhEz7HbeMAwjGhI/hp46QdTlqqJS5z/tzDbAzk/+4sx2bugJBI3LVXxgbX4cyR1tccoEUcMw4g8LooZhGFFgQdQwDCNCcsuyTwuihmHEjsSPobHPJxouLuVkI5XBfeHGLqx78yrmPzf4+L5SRQsw4eF+/PC/oUx4uB8lixYA4IzkUkwfNYDdn4zkxgubHS+fXLYokx+9kIUvDGHB84MZ2bdJWD7UrV2DVs0a06ZlM9q3bRXWudmRqJLJ69evp3vXTjRrXJ/mTRrwzNOjA7MNsHv3bi69eABNG9ajWaP6zJn9XSB2TxnJZMkda+cTLoi6lJONVAZ37JfL6XvPiYqJtw5syfTF62l09etMX7yeWwe0AGDXvsPc8sLXPPXBwhPKp6Qe486XZtD82nGcc/O7XNO7MXWrhqcmOemLacyZv4hZs+eF/RkyI5Elk/Pmzctj/xrFou+X8fXM2fz3hWcD9f22m2+kW/fuLP5xOXMWLKZO3XqB2D1lJJOxIBoTXMrJRiqDO+vHTezcd/iEfb3b1mTcl8sBGPflcvq0qwXAtj2HWPDzVo6mHjuh/JZdB1n8i5fZf/+ho/z06y4qly0SyccIlESWTK5UqRLNmjcHoFixYtStW49NmzYGYnvPnj3MnPkNlw/3kpPlz5+fkiVLBmL7VJJMNo2lGJGamkqbFk2pVrk8nbt2i0vp4fIlC7Nll6dzv2XXQcqXLBzyudXKF6NprXLM++m3kM8REfr06s6ZbVry8ksvhu1vZmQksbtxYzCBKCfsp7Fu7VoWL15Eq9bBXCtr16yhbNlyXHPVFbRt1Zz/u+YqDhw4EIhtcHeN51R7h4r1RLNARKqKyFciskxElorIDUHZTkQ52VAncBcpmI+3/n4et734DfsOHQnZ/pdfzeC7uQv4+NOJvPj8c8yc8U2kruY69u/fz6CBF/H4qKcoXrx4IDZTUlNYvGghV11zLbPnLaRIkSL8O8DxxUS8xsMlnAB6SgZRIAW4RVXrA22BkSJSP8gK4ll6eOvug1Qs5fU+K5YqzLY9h7I9J29SHt76ey/emb6CT779Jaz60mRvy5cvT5++/Zg/b274TmdAoksmHz16lEEDL+LiQYPpd8GFgdmtUiWZKsnJtPZ7thdc2J/FixcFZj+NXC2ZjPVEs0RVN6vqQv/1PmA5EPW3lSjSw5/NXs2Qrt6DhiFd6zFh9upszvCe8q9Yv5OnPwrvx3jgwAH27dt3/PXUL7+gfoPwZhhkRiJLJqsq1159JXXq1uOGm24OxGYaFStWJDm5KitXrADgq2lTqVcvmAdLJpmcWEE0R+aJ+qqfzYA5GRwbAYwAqFqtWra2tmzezNVXDCM1NZVjeoyL+g8MTE526JBBzPh6Otu3b6dW9WTuufcBLr8ie1WTMbd3p0PjZMoWL8iq16/gwXGz+fd7Cxh3V0+GnduAX7fuZcijkwCoUKows0ZfQrHC+Tl2TLm+XzOaXTOORjXKMLhLPX5Ys53Z//FEAe4b8y1T5q/Ltv6tv/3GJQO8XlZKSgoDLxnEud17RNESf5BeYjc1NZVhl1/hTDI5aPvfzprFm2+MpWHDRrRp0RSABx56hB49ewVif9STTzN82BCOHjlC9Ro1+e9LrwRi1+U17vr7DJv4jY0h41QyGUBEigJfAw+r6odZlXUpmewSS0CSu7AEJH/GhWRygQq1tcrg0OfurnnyvFNPMllE8gEfAG9kF0ANwzjFsCxOWSNe67wMLFfVJ1zVYxhGYiJALoihTp/OtwcuAzqLyGJ/C2YwyjCMXICQJ0/oW7ziUjJ5Jrli2NgwDFfkhtv5hFyxZBhGLkC82/lQt2zNidzkL+z5UUTeEpGCIlJDROaIyCoReUdE8vtlC/jvV/nHq0f6MSyIGoYREwQCu50XkSrAX4GWqtoQSAIuAf4JPKmqpwO7gLQ5i1cCu/z9T/rlIsKCqGEYMSPInije8GQhEckLFAY2A52B9/3jY4B+/uu+/nv8410kwrEFC6KGYcSMoFYsqepG4N/Ar3jBcw+wANitqil+sQ38sWqyCrDePzfFL18mks9gQdQwjNgQ/phoWRGZn24bcdyUSCm83mUNoDJQBAhm2V42mDyIYRgxwZsnGtYd9PYsVix1Bdao6jY8ux/iTbMsKSJ5/d5mMpCW928jUBXY4N/+lwB2hP8pLIgGwq7xf3Vqv1T/4PKDnsyu90dkX8hIGFwtWXVjNdDEIr8CbUWkMHAI6ALMB74C+gNvA8OAtAzU4/333/nHp2mEjWdB1DCMmBFUDFXVOSLyPrAQLw3nIuBF4DPgbRF5yN/3sn/Ky8BYEVkF7MR7kh8RFkQNw4gNQqArkVT1PuBkVb/VQOsMyh4GBgRRrwVRwzBiQgRjonFJQj6ddyn5Gu/yvSN7N2T+6P4seLo/1/fxEi83ql6a6Y/1Zd7o/rz/9+4UK5QPgM5NqjBr1AXMG92fWaMu4JxGlSOq02V7u5Y1dum7K8lkiFy+OztWrlhBm5bNjm8VypTgmaefCrSOcAh4nmhMSLgg6lLyNd7le+tXK8XwbnXpcNtHtL7xA3q2rEbNisV5fuTZ3D12Lq1ueJ/xs9dy0wWeZv2OvYfp/9AUWt3wPlePns4rN3YK22/XErsuZY1d++5KMhkil+/OjjPq1GHO/EXMmb+Ib+fMp1Dhwpzf94LA6wmV3JDZPuGCqEvJ13iX762bXJJ5P2/l0JFUUo8pM5Zupl+7GpxeuSQzl24GYNqSDfRrVwOAJWt2sNlXHF326y4K5k8if97wvnLXErsuZY1d+u5SMhkil+8Oh6+mTaVmzVpUO+00p/VkhfVEY4BLydd4l+9d+usu2terSOliBSiUP4kezauRXLYIy9fvpE8b74dw4Zk1Sc5Ar/6CdjVYvHo7R1KO/elYVuSkxG7QssYufXctmZwTvPfu2wy4OOKH0tEj1hPNEj+DylwRWeJnVnnAVV2JRDTyvSs27GbUR0v49P5ejL+vF0vW7CD1mHLNf75mRM8GzBp1AUUL5ePI0RMDZb2qpXhoWBuuf35GkB8lUFzIGrvEtWSya44cOcLECZ9y4UWBPKCOiLSkzIneE3X5dP53oLOq7vdlQmaKyCRVnR2NUZeSr4kg3zvmyxWM+dJTmHxgSCs27jjAyo176HP/RABOr1yCni3+EPyrUqYI79zZjaue+oo1W/aFXV9OSOy6kjV26XtGksn/fjziREA5zpTJk2jarDkVKlSIoRfx3cMMFZeSyaqq+/23+fwt6oUPLiVfE0G+t1yJggBULVuEvm1r8M43q47vE4E7BzTjf1OWA1CiSH4+vLsH94ydy3c//RZRfa4ldl3KGrv03aVkck7w3jsxvpX3sZ5oNohIEl4mldOBZ1X1T5LJ4eJS8jUR5HvfuqMbpYsV5GjKMW58cSZ7DhxhZO+GXNOzPgCfzF7L61O9H/a1vRpQq1Jx7rq4OXdd7D286XP/RLbtORxyfa4ldl3KGrv23ZVkMkQu3x0KBw4cYNrUL/jPcy8EYi9iAp5sHyucSyYDiEhJ4CPgL6r640nH0uvOt1j5S/Y666catnY+Z0lkyWRXvrdv24qFAUsmF6taV5ve+FLI5Wfe2iEuJZNz5Om8qu7GSwTwp9RUqvqiqrZU1ZblypbLCXcMw4gT7Ol8FohIOb8HiogUAroBP7mqzzCMxMPGRLOmEjDGHxfNA7yrqhMc1mcYRoIRzz3MUHEpmfw90MyVfcMwEpw472GGimVxMgwjJkgumSdqQdQwjJiRC2KoBVHDMGJHnlwQRS2IGoYRM3JBDLUgahhGbBCBpFywYsmCqGEYMcMeLBk5gsulmeWGjHFme9u4Yc5suySRf9g5sIo7UBK4qY+TaRAVkf+QRdYlVXUrtm4YRq5G8KY5JTpZ9UTn55gXhmGckuSCIdHMg6iqnnCfJyKFVfWge5cMwzgliPPEIqGSbQISEWknIsvwk4eISBMRec65Z4Zh5HpyQwKSULI4PQV0B3YAqOoS4GyXTmWFKz3uNFzqlIMn49u2ZTMu7Ns7ULtB+P1/Pesx5/Hzmft4X67r6WVpv6t/E1Y8N4BZj/Vh1mN9OLepJ68xsH2N4/tmPdaHPW8OpdFppWLme0YcPnyYs9q1pnXzJjRv0oAHH7gvMNvr16+ne9dONGtcn+ZNGvDM06Pj2va1I67gtOQKtGzW6Pi+f9x/D61bNKFtq2b06dWdzZs2RV1POAjeZPtQt3glpFR4qrr+pF2pDnwJCVd63OBepxzgmadHUydgGYkg/K6XXJLLO9em498/o90d4+nRPJmaFYoB8OzEZbS/81Pa3/kpny/21DLfnbXm+L6rn53B2m37+WHdrpj4nhkFChRg8hfTmLtwCXPmL+bzKZOZMzsqia/j5M2bl8f+NYpF3y/j65mz+e8LzwbmtwvbQy67nI8/nXTCvhtvvo25C5Ywe94ievY6j0cf/kdUdUTCqdITXS8iZwIqIvlE5FZguWO/MsWlHrdrjfUNGzYwedJnDL/iqsBsQjB+16lSgvmrth/XtJ+5/DfOb10t+xOBAe1r8MG3ayJx3WmbiwhFixYFPDG8lKNHAxuDq1SpEs2ae5IrxYoVo27demzaFIwcswvbZ3U4m9KlTvzdpFdVPXDwQEzGJ0+VpMzXAiOBKsAmoKn/PtfhWmP9tltu5OFH/0WePMHmwg7C7+Xrd3Nm3fKULupp2ndvWoUqZTz9+hHd6/LdP/vw3DVnUrJI/j+de2G7Grw3K7Ig6rrNU1NTadOiKdUql6dz1260bhOMpn161q1dy+LFi2jVOrFsA9x/7985o1Y13nnrTe6+L2d7omkrlkLd4pVsf82qul1VB6tqBVUtp6pDVHVHqBWISJKILBKRUzoh88TPJlC+XHmat2gRa1cyZMWmPTw5/kc+/ls3PrqrG9+v20XqMeWlL1bQ+K8fcuadn7Jl9yEeGXKixE3L08ty6PcUlm/YHSPPsyYpKYk5Cxazau0G5s+by9Iff8z+pDDYv38/gwZexOOjnjqhZxfvttO4/x8Ps/KXX7l40KX89/lnnNSRFRLGFq+E8nS+poh8KiLbRGSriHwiIjXDqOMGYnj7Hw4udcq/+3YWEyaMp87p1Rk6+BKmfzWN4UOHBGI7KL9f/2oVZ/9tAj0emMzuA7+zavNetu05zDFVVOG1aStpcXrZE8656MwavB/hrXyQvmdHyZIlOadjJz7/PLjx9KNHjzJo4EVcPGgw/S64MDC7rm1nxCWXDObjjz50Xs/JnCq3828C7+LJfVQG3gPeCsW4iCQD5wGhS/rFEJc65Q8+/Ci/rN3AilVref2Nt+nYqTOvvj4uENtB+V22uKdfn1ymCOe3Oo33Zq2mQslCx4/3aXUay9b/0eMUgQvbVo8qiLps823btrF7t+fvoUOHmPrlF9SpUzcQ26rKtVdfSZ269bjhppsDsZkTttOz6uefj7+e8OkngbVNqHhP50Pf4pVQ1s4XVtWx6d6PE5HbQrT/FHA7UCyzAidJJmdr0KUet2udclcE5fcbN3ekdNECHE09xs2vzmbPwaM8PrwNjU8rjary67YD/PWl746Xb1+vAht3HGDt1v0x9z0jtmzezNVXDCM1NZVjeoyL+g+k13nBTC37dtYs3nxjLA0bNqJNi6YAPPDQI/To2SsubQ+77FJmfDOdHdu3U7tmVe6+536mTJ7EypUryJMnD9WqncbTzzwfte9hEXAP0xfGfAloiLdk/QpgBfAOUB1YCwxU1V3iVTwa6AUcBC5X1YUR1ZuZTrWIpD3KuwPYBbztO3YxUEpV78rmA/UGeqnqdSLSEbhVVbO8glu0aKmz5thq05zEEpDkLh6FDEIAACAASURBVI4dc5OB5Kx2wevOl6nZQHs9+GbI5ccNaZql7ryIjAFmqOpLIpIfKAz8Ddipqo+JyJ14sesOEekF/AUviLYBRqtqRE/vsuqJLsALmmkNd026YwpkGUSB9sD5vrMFgeIiMk5VgxkINAwj4QmqJyoiJfAWAV0OoKpHgCMi0hfo6BcbA0zH6xj2BV5Xrxc5W0RKikglVd0cbt1ZrZ2vEa6xk86/Cz/QpuuJWgA1DAP4Y0w0IGoA24BXRaQJXifwBqBCusC4Bajgv64CpF9EtMHfF1wQTY+INATq4/UoAVDV18OtzDAMIz1h9kTLikj68b4XVfVF/3VeoDnwF1WdIyKjgTvTn6yqKiKBj3dkG0RF5D687nB9YCLQE5gJhBxEVXU6XjfaMAwD8CfbhxdEt2cxJroB2KCqc/z37+MF0d/SbtNFpBKw1T++Eaia7vxkf1/YhDLFqT/QBdiiqsOBJkCJSCozDMNIT1Br51V1C94S9Tr+ri7AMmA8kPaEcxiQtqZ4PDBUPNoCeyIZD4XQbucPqeoxEUkRkeJ4kbxqdicZhmFkR8CT6P8CvOE/mV8NDMfrKL4rIlcC64CBftmJeE/mV+FNcRoeaaWhBNH5/vyr/+EN1u4Hvsv6FMMwjOwJMoaq6mIgo9v9LhmUVQLKAZJtEFXV6/yXL4jIZKC4qn4fROWGYZy6CPGdJzRUshKqa57VsUhn9xuGYQAQ53lCQyWrnuioLI4p0DlgX4xMcLUKBdyuKjrjpvHObK98Mpj19bmNPI4WmbuKdfGcWCRUspps3yknHTEM49Qj2My6sSGkyfaGYRhBI+TynqhhGIZr4jnFXahYEDUMIyakyYMkOqFkthcRGSIi9/rvq4lIa/euZYxLGdxEk2POKRncINrlyk41+fJvHfniro785/LmFMjrXXq39a7L9Hs6M/XvnRh+zok5bxpXK8nqp3rTq2mlqPyPZ5nqjHApx5yGqzYJl9yQlDmUcd3ngHbAIP/9PuBZZx5lg0sZ3ESTY84pGdxo26VCiYIMP6cG5z3+Dd0enU6SCH1aVGFAm6pULlWITg9No8vDXzF+wR9Ll/MI3NW3Ht/8tC1q/+NVpjozXMoxp+GiTSLhVJFMbqOqI4HDAKq6C/iz5GMO4VIGN9HkmHNKBjeIdsmbJw8F8yWRlEcolD+J3/Yc5rIO1Xlq0krS8oLv2H/kePnh59Rk0uLN7Nj/e1T1xrNMdWa4lGMGd20SLl4qPAl5i1dCCaJHRSQJb24oIlIOOObUq2zICRncoHEtDZyeWMrgZsRvew7z4tRVzP5HN+Y/dC57D6Uw46dtnFa2CH2aV2bCbWcz5v/aUL2cJ9FcoURBujeuyNiZa6OuO55lqkPBhWSyqzaJhDxhbPFKKL49DXwElBeRh/HS4D0SinERWSsiP4jI4pPyAEaFaxncRCfWMrgnU6JQPro1rkj7+7+k1d2fU7hAEhe0TCZ/3jz8nnKM3o9/w1vf/sq/B3t6Qvdf1JBHxy8nE+WakIl3merscCGZHG9tkhtu50NZO/+GiCzAW8QvQD9VDUcCuZOqbo/UwaxIL4PboKGbh0FBkVPSwOm55JLBXND3PO6+9wGn9WTHWXXKsn7HQXb6t+uTl2ymRc1SbN59iMlLNh/flxZEG1UrwTOXez/y0kXz06l+BVKOKZ9/vyWsetNkqidPnsjvhw+zd+9ehg8dEojKquvv05Vksss2CReJ89v0UAnl6Xw1vFRRn+Ll4Dvg74sJLmVwXeJSGjg9sZbBzYiNuw7RvHopCuZLAqD9GeVYtWU/n3+/hXa1ywDQ9vQyrPFVQ8+6fyrt7/+S9vd/ycTFm7j73e/DDqCQGDLVGeFSMtllm0TCKdETBT7jD8G6gnhaJiuAUHRtFfjcT8n/33Sp/I8TrmSySxncRJNjzikZ3GjbZfG63UxcvJmJd5xNaqqydMMe3vx2HQXz5WH0sBZc1akWB35P4fa3lkTta07hUurZpRxzvBHPU5dCJVPJ5ExP8LI7Xaeq2T7aE5EqqrpRRMoDX+Dpn3yTWXmTTM4YlwlIXCWsAEtAkpto36YlCwKWTK5yRiO99rmPQi5/b7faWUomx4qwH3r5KfBCelSoqhv9/7fiPZyK2SR9wzDijDAm2sdzjzUUobr0gzJ58BT1sl0GIyJFgDyqus9/fS4Q+/k2hmHEDeIsyV7OEcqYaLF0r1Pwxkg/COG8CsBH/mTvvMCbqupmOZBhGAlHwLrzMSPLIOpPsi+mqreGa1hVV+MpgxqGYWRIrg6iIpJXVVNEpH1OOmQYxqlDbs8nOhdv/HOxiIwH3gMOpB1U1Q8d+2YYRi7mlLid9ykI7MDTVEqbL6qABVHDMCInzifRh0pWQbS8/2T+R/4Inmm4m7hoGMYpQ25Y9plVEE0CipKx0J8FUcMwouJUuJ3frKo2rzMOcLmqyCU/jerjzHblK950ZnvTK5c6s22kR0jK5T3RxP90hmHELZ7aZ6y9iJ6sgmiXHPPCMIxTjzhfzhkqmQZRVd2Zk44YhnHqkdsfLBmGYTgjt9zOx7N0SYa4lEx2adu1DK5ruecgJXYzknreuXMnvXueS+P6Z9C757ns2rUrZHvXnFuHWY/04ttHenFt9zoA/O2ixsx4qCdfP9iTD27rRMWShQCoXak4U+49l80vX8z1PaNLWO1KMjmNRJN6joRTRagurnApmezStmsZXJdyzxCsxG5GUs+jHn+Mjp078/2ylXTs3JlRj4f2465XpQRDO9ai6/1T6HD3JM5tWoUa5Yvyn8+W0eHuSZxzzySmLN7Ibf28Py679v/OnWPn88ykcBRu/oxLyeQ0Ek3qORJyQ2b7hAuiLiWTXdp2LYPrUu45aIndjKSeP/t0PIOHDANg8JBhTBgfmvzwGZWLs+CXHRw6kkrqMeXbn7bSu2VV9h1OOV6mcIG8x0Xvtu/7nUVrdpKSGt1UZ5eSyZCYUs/hIpw6ap9xh0vJ5JyQY3Yhg+uSnJDY3br1NypVqgRAxYoV2br1t5DOW75xD23rlKNU0fwUyp9EtyaVqVKmMAB/79+YH57sy4Azq/Poh98H6q9ryeREl3oOCfE6LqFuIZkUSRKRRSIywX9fQ0TmiMgqEXlHRPL7+wv471f5x6tH+jGcBlERKSki74vITyKyXETaBWHXpWSyazlmFzK4LomFxG44P5qVm/by9IRlfHBbZ967tRM/rNt1XE7l4fe/p9FNn/Det2u5uusZLl0OlHiTNXaJhLGFyA1A+rGafwJPqurpwC4gTRzsSmCXv/9Jv1xEuO6JjgYmq2pdvNyi0Q1EnUR6yeSgcWHblQyuS9IkduucXp2hgy9h+lfTGD50SOD1lC9fgc2bPfnkzZs3U65c+ZDPHffNajrfN5nej3zJ7gNHWLVl3wnH3/tuLX1aVc345AhxKZnsss1jId2dGQIkiYS8ZWtPJBk4D3jJfy94iZPe94uMAfr5r/v67/GPd5EIx+6cBVERKQGcDbwMoKpHVHV3tHZdSia7tO1SBtclOSWx26t3H94Y513Tb4wbw3l9QheiK1usAABVyhSmd8tk3v9uLTUr/CHI0Kt5Mj9v2huovy4lkxNV6jkSwnywVFZE5qfbRpxk7ingduCY/74MsFtV0wbINwBpfzGqAOsB/ON7/PJh43KeaA1gG/CqiDQBFgA3qOqB9IXiSTLZpW3XMrgu5Z6DJiOp51tuu5PLLr2Y1199harVTmPsm++EbG/MXztQumgBjqYe4/bX57P34FH+c2UbTq9UnGPHlPU7DnLLa3MBKF+iINMe6EGxQvk4dky5tntd2t054YQHUaHgUjLZJfHld+jDNj7bM1P7FJHewFZVXSAiHYPwLlTClkwO2bBIS2A20F5V54jIaGCvqt6T2TkmmZy7cCn1nHzVW85sWwKSP+NCMrlW/Sb6yBsTQy5/SfPkTCWTReRR4DI8HbiCQHE8heHuQEVfpaMdcL+qdheRKf7r70QkL7AFKKcRBESXY6IbgA2qOsd//z5epnzDMAwguKfzqnqXqiaranXgEmCaqg4GvgL6+8WGAWnzucb77/GPT4skgILDIKqqW4D1IlLH39UFiN2sXsMw4g4HT+dP5g7gZhFZhTfm+bK//2WgjL//ZuDOSCtwvXb+L8Ab/tys1cBwx/UZhpEoiBuhOlWdDkz3X68GWmdQ5jAwIIj6nAZRVV0MZDiGYRjGqU3aiqVEx7I4GYYRM3K7ZLJhGIZTcnVSZsMwDJd4t/OJH0UtiBqGETNywd28BVHDMGKFINYTNcDtyhxIXMlkl367XFVUqt+zzmwD7Pp4pDPbrlYgurrCrSdqGIYRITYmahiGEQ1xLvsRKhZEDcOIGRZEDcMwoiA3PFhKuFVXLqWBg5ZMDloaODNcSj2DW4ld11LPkfj+wg2dWTduOPOfveT4vlJFCzDhwfP54cXBTHjwfEoWKXD82KgRHfjxxSHM/c/FNK1V9gRbxQrlY9Vrw3jy2g5h+e2yXerWrkGrZo1p07IZ7du2Ctx+qAjeZPtQt3gl4YKoS2ngoCWTg5QGzgqXUs+uJXZdfp+R+j72y+X0ve/TE/bdOqA505dsoNGIN5i+ZAO3DvCyOnZveRq1Kpeg4YhxXP/MdJ6+ruMJ5913WRtm/rgpbN9dS2BP+mIac+YvYtbsec7qCAXTnY8BLqWBg5ZMDlIaOCtcSj27lth1+X1G6vuspZvZue/3E/b1blODcVN/AmDc1J/o07bG8f1vTlsBwNwVv1GiSH4qlvLURpvVKkf5koX5ctF6wsVlu8QTEsa/eCXhgqhrXEsmRyoNnB2u/I4rid0wCdL38iULs2XXQQC27DpI+ZJeoKxcpggbtu8/Xm7jjgNULlMEEXjsqvbc9fKsKD6BG0SEPr26c2ablrz80oux8wO7nc8SEakjIovTbXtF5EZX9QWFa8nk9IQjDZwdOem3AZrN9PNrzmvElPnr2LjjQJblYsGXX83gu7kL+PjTibz4/HPMnPFNjDwJpx8av1HU2dN5VV0BNAUQkSRgI57mSUKQXjK5QcPgBvfTpIErVaoUtjRwKATtdzxJ7IZLkL5v3X2QiqW83mjFUoXZtvsQAJt2HCC5bNHj5aqUKcKmHQdoU7ci7etXYkSvhhQpmI/8+ZLYf+go94wJZqw6GtLaoHz58vTp24/58+ZyVoezc96RXDJPNKdu57sAv6jquhyqLyJcSianEY00cGa49DveJHbDIUjfP5uzliFdvDYd0qUuE+as8fev4dLOngJO6zoV2HvwCFt2HWT4v7/gjCtep+6VY7nrlW95c9pPcRFADxw4wL59+46/nvrlF9Rv4GZmRCjkgDyIc3IqiF4CZCjPKCIj0nSkt23flq2hoUMG0bFDO1auWEGt6sm89srL2Z4TKls2b6ZH1060ataYs9q1okvXblFJJg+77FI6nXMmP69cQe2aVRnz6svcctudTPvySxrXP4Ovpk7lltsilnZx5nd60kvsNm1Uj4sGDAxUYtfl9xmp72Nu68b0f1/EGVVKsuq1YQzrVo9/v7+Azs2q8sOLg+nUNJl/v7cQgMnz17Fmy16W/m8Iz/6lEzc893Ugvrtql62//UbXjh1o06IpZ5/Zhh49e3Fu9x6B2A4Xb0w08Z/OO5NMPl6Bp6+0CWigqlk+RUlUyWRLQJK7sAQkf6Z921YsDFgyuV6jZvrqR1+FXL5d7VKZSibHkpxYsdQTWJhdADUM4xQkF/QPciKIDiKTW3nDME5t4vk2PVScjomKSBGgG/Chy3oMw0hMcsODJdeSyQeAMi7rMAwjgYnn6BgilsXJMIyY4PUwEz+KWhA1DCM25JLJ9hZEDcOIGbkghloQNQwjhuSCKGpB1DCMGBHfiUVCxYKoYRgxw8ZEDcD9heByWalL34NK85fTuFyWCVBx2DhntreMGeLErotvMt7nf4aKBVHDMGJGov6hTY9ltjcMI2aIhL5lbUeqishXIrJMRJaKyA3+/tIi8oWI/Oz/X8rfLyLytIisEpHvRaR5pJ/BgqhhGDEjwGWfKcAtqlofaAuMFJH6wJ3AVFWtDUz134OXGKm2v40Ano/0MyRcEI1Hid1QCVqqNqckmcGdzK5LuWfX1wp42lZtWzbjwr6R5W+9tnsdvn2sN9/9szf/1+PERNrX96rH7jeGULqoJ8884MzqzHr0PGY9dh5T7utOw2olw67Ptbx2WIQTQbOJoqq6WVUX+q/3AcuBKkBfYIxfbAzQz3/dF3hdPWYDJUWkUiQfI+GCaDxK7IZDkFK1OSXJnIYLmV2Xcs+uZYcBnnl6NHXq1Yvo3HrJJRjaqTZd7p3EWXd9RvdmVahRwZMaqVK6MJ0aVWJ9OhG8ddv20+vBL2h/52c8/vEPPHVl27DrdNnekRCmxlLZtATu/jYiQ5si1YFmwByggqpu9g9tASr4r6sA6WVYN/j7wibhgmg8SuzGipySZHaJS7ln17LDGzZsYPKkzxh+xVURnX9G5RIs+GU7h46kknpMmbV8K31aVQPgkctacN9bC0mfY3nuz9vZc/AIAPN+3k7l0oXDrtNle4ftC2GPiW5X1Zbptj9JlYpIUeAD4EZV3Zv+mHoZqwOf6pJwQdQlruWBc0Kq1pUks0vfXctUu+K2W27k4Uf/RZ48kf2Mlm/YTbs65SlVND+F8ifRrWllkksXpleLZDbvPMSPv+7O9NzLOtbiyyWbIqo3nto7yFR4IpIPL4C+oapp6Td/S7tN9//f6u/fCFRNd3qyvy9sXOcTvcl/UvajiLwlIgVd1hfv5LRUbZCSzC59T0S554mfTaB8ufI0b9EiYhsrN+1l9KdL+ejOLnxwR2d+WLeL/PmSuPn8hjzy/pJMz+tQvwKXdTyd+95eGFG9cdXeAUVR8S70l4HlqvpEukPjgWH+62HAJ+n2D/Wf0rcF9qS77Q8Ll7rzVYC/Ai1VtSGQhCdYF7e4lgfOSKo2aNIkmYFAJZlzwvf0cs/xznffzmLChPHUOb06QwdfwvSvpjF8aPgT3cd+/Qsd755Erwe/YPeBI/y0YTenlSvKzEfP4/un+lG5dGG+frgX5Ut4/Y8GVUvy9FVtufSJ6ezafySqzxAP7R2g7nx74DKgs4gs9rdewGNANxH5GejqvweYCKwGVgH/A66L9DO4vp3PCxQSkbxAYTzBurjFpTxwTknVupBkdul7TshUu+DBhx/ll7UbWLFqLa+/8TYdO3Xm1dfDX4lUtrj35D25TGH6tKrKWzNWU/u692l848c0vvFjNu08yDl/n8jWPYdJLlOYsTeewzXPz+KXLfsi8jve2juPhL5lharOVFVR1caq2tTfJqrqDlXtoqq1VbWrqu70y6uqjlTVWqraSFUjVsh0tmJJVTeKyL+BX4FDwOeq+vnJ5fwnbCMAqlarlq3doUMGMePr6Wzfvp1a1ZO5594HuPyKKwPxOb3EbmpqKsMuvyIweeCtv/3GJQMuBCAlJYWBlwyKWqp22GWXMuOb6ezYvp3aNaty9z33c8ttd3LZpRfz+quvULXaaYx985249D2NLZs3c/UVw0hNTeWYHuOi/gMDk3t2ea0Exes3nEPpYvlJSVFufW0eew4ezbTs7Rc0pnSx/Iwa3hqAlFSl0z2TMi2fES7bOyISf8GSO8lkf2XAB8DFwG7gPeB9Vc30z3WiSia7lp12ad7Wzuc8ibh2vn2bliwIWDK5UZPm+uHns0Iuf0bFwnEpmezydr4rsEZVt6nqUTyxujMd1mcYRiIRxvSmeP577DKI/gq0FZHC/pOzLnirCAzDMIBgpzjFCpdjonNE5H1gId661kWAm8mRhmEkJvEcHUPEtWTyfUAMF+cahhG/WGZ7wzCMqIjnsc5QsSBqGEZMiPexzlCxIGoYRuzIBVHUgqhhGDEjTy64n7cgahhGzEj8EGpB1DCMWBHnk+hDxYJoALhe3ujSfKLKMbvE9TJeV0szAUoNfNmJ3d9Xb3diNzf0RS2IGoYRE9Iy2yc6FkQNw4gZuSCGWhA1DCN25IaeaEJqLLmSNV6/fj3du3aiWeP6NG/SgGeeHh2YbXArxwzRy/dmxrP/GU3LZo1o2bQhzzz9VKC2E7XNV65YQZuWzY5vFcqUCLRtgvB75HkNmP/UhSx46kKu7+3lxW1cvTRfP9aH2aP6MfNf59Py9LIAdGhQkS1jL2P2qH7MHtWPuwY0DeyzZEWAme1jRsL1RNNkjT+b9AVVkpM5q20revc+n3r160dtO2/evDz2r1E0a96cffv2cWabFnTp2i0Q2y79TiNNvnff3r3ZFw6RpUt/5NVXXuKbWXPInz8/fXv3pGev3tQ6/fRA7Cdqm59Rpw5z5i86Xk+t6smc3/eCqO2m2YvW7/rVSjG8Wx063P4JR1KOMf6e7kyc/ysPD23Nw+8s4vNFG+jePJmHh7am+70TAZi1fAsXPfJFIJ8hZOI3NoZMwvVEXcoaV6pUiWbNmwNQrFgx6tatx6ZNwah9upZjjla+NzNW/LScVq1bU7hwYfLmzUuHs8/mk48/zP7EEEnkNk/jq2lTqVmzFtVOOy0Qe0H4XbdKCeat3HpcjnnGsi30a1sdVaV44XwAlCicn807Dwbic6TkhlR4CRdEXcsap7Fu7VoWL15Eq9bByMm69jta+d7MqF+/Id/OnMmOHTs4ePAgUyZPYmM6Mb8gSbQ2T+O9d99mwMXBaTAG4ffSX3fRvn5FShctQKH8SfRoXpXkskW47ZXZPDK0NT+/eDGPDmvNvW/8oSTRpk555jzRj4/vPpd6VUsG9nkyQ8RbsRTqFq+4lky+wZdLXioiN7qsK0j279/PoIEX8fiopyhevHis3cmWIOR7M6NuvXrcfOvtnH9ed/r16Unjxk3Ik5QUeD2J1uZpHDlyhIkTPuXCiwbE2pUTWLFxD6M++p5P7+vB+Ht6sGTNDlKPKSN61OP2V+dQe8Q73P7qHJ6/7iwAFq/eQZ1r3qHNzR/z/MRlvHtH15xxNBd0RV1KJjcErgZaA02A3iIS9UCaa1njo0ePMmjgRVw8aDD9LrgwMLsu/Q5Kvjczhg2/klmz5/P51K8pWaoUtWufEZhtSMw2T2PK5Ek0bdacChUqBGYzKL/HTF1J+9s+ods9n7H7wO/8vGkvgzvW5uPZawH44Ns1tKxdDoB9h45y4HAKAFMWbiBf3jyUKVYg+g+TDbkghjrtidYD5qjqQVVNAb4Gov6FuJQ1VlWuvfpK6tStxw033RyIzTRc+h2UfG9mbN26FYD1v/7K+I8/YuAllwZmO1HbPI333gn2Vh6C87ucr1VftWwR+rapzjvf/MLmXQfp0KAiAB0bVWLVZu8hZIWShf6o//Sy5BFhx77fA/g0WZMbNJZcPp3/EXhYRMrgSSb3AqKW8nQpa/ztrFm8+cZYGjZsRJsW3hSPBx56hB49e0Vt26Xfrhl8SX927thB3nz5eGL0M5QsGdx4WSK3+YEDB5g29Qv+89wLgdmE4Px+67YulC5WgKOpx7jxf9+y5+ARRj43k8evbEveJOH3I6lc//xMAC5oV52ru9cj5dgxDh9JZegTXwX6mTImvqcuhYozyWQAEbkSuA44ACwFflfVG08qk153vsXKX9Y588f4My7XzufJk5g/ENdr513mWnC2dv6rBzm2a22gjjdr3lKnzZwTcvnSRfKecpLJqOrLqtpCVc8GdgErMyjzoqq2VNWW5cqWc+mOYRhG4DidbC8i5VV1q4hUwxsPbeuyPsMwEot4HusMFdcrlj7wx0SPAiNVdbfj+gzDSCByw5ioa8nkDi7tG4aRuHiT7WPtRfQk3Np5wzByERZEDcMwIsdu5w3DMKIgNzxYSrgEJIZh5B6CXPYpIj1EZIWIrBKROx25/CcsiBqGETsCiqIikgQ8C/QE6gODRCS4ZL1ZYEHUMIyYEWBm+9bAKlVdrapHgLeBvs4/AHE2Jrpw4YLthfJJqOs+ywKudFxd2nZt32znHtuu7YdjO5iM0+lYtHDBlML5pWwYpxQUkfT5N15U1Rf911WA9IluNwDBJKbNhrgKoqoa8rpPEZnvah2tS9uu7Zvt3GPbtX3XvmeHqvaIVd1BYrfzhmHkBjYCVdO9T/b3OceCqGEYuYF5QG0RqSEi+YFLgPE5UXFc3c6HyYvZF4lL267tm+3cY9u1fde+5xiqmiIi1wNTgCTgFVVdmhN1O80nahiGkdux23nDMIwosCBqGIYRBRZEjZAQl5oWjhCRIg5tV0zENjGCJ6GCqIjUEZF2IpLPX+YVtP3gBdU9u6eLSEsRCVyDVkQaiMg5fvLroG2fJSKXAaiqBh00RKSPiNwQpM10tvsC/xSR8g5sdwc+4sQpNUHZbisil/n/5w/Ydm3/Oszj6lo/FUmYICoiFwKfAA8BLwMjRaR4QLbPAFDV1KAvLhHpDXwIPA68llZXQLZ7Am8BNwGvi0jFgOzmEZGiwH+Bu0TkWjgeSAO5ZkTkXOBBYFkQ9k6yfQ7wT+ATVd0asO1zfduVgFsCtn0+3hPzrsCtBLhKSET6Ae8DdwFPANe47KmfSiREEBWRfMDFwJWq2gUvmFYF7og2kPpBbrGIvAnBBlIRORMveA5T1U54Yn2BZJcRkY7AaOAqVe0HHAEaBmFbVY+p6n5gDN4frDNF5Ka0Y9Ha99tlLDBCVb8QkRIicpqIFI7Wtk8L4CXfdmUR6SYibUSkRDRGRaQr8BwwGKgN1BORswPwF/9OYiRwqaoOA/YCTUWkvIgUDMD2NcAgVb0I+B4YDtwsIsWidP2UJyGCqE9xvAsXvFupCUA+4NJIbzP9v8TXAzcCR0RkHATeI/2nqi7yX98HlA7otv434BpVnev3QNsA14vIf0Wkf0C33il4f6zGAK1F5AkReVQ8orl2duDpblXyf+AfA8/j9dSD8D0l3ev3gSvwvudnRaRUFHaTgKH+/MMiwAqgAQQyZpwCFALq+h2DjsBQcSuZwAAABjJJREFU4Cng7ih7jSlAUaAigKq+AqzFWzvfOwq7Bnga24mwAd3wViB08N8nAZcC4/Dnu0ZotzLeBVYW7wc3LkCfk4Di6V4nA4uAcv6+MgHV83fgbv/15XgZbMoFYLcWcKf/+hbgIPBsQD43AVbjJYq4Gu8P+hV4wxOlo7TdCC/AvQ0M9/fVBF4Augfgex7//x7AFqBRQG3SH1gAzAbu8fd1Bl4DmkRp+1r/t3IZ8LD/+hrg5SB8P5W3ROqJzgA+By4TkbNVNVVV38QLgk0iNaqqm1R1v6pux7uoCqX1SEWkuYjUjcJ2qqru9d8KsBvYqarbRGQw8JCIFIrUfrp6HlbVh/zXr+H12oN46HEIqCMiV+P9CB8DqonINdEaVtUleL2gx1T1f+oNIbwClAKqRWn7B7wxxTZADX/farw/ZCEnucnC/jH//8l4Y5i9A+ido6rv442HzsD7Y4uqTgOKEf346FvAJKATUEhVh6jqf4EKQT1bOFVJmGWfqnpYRN4AFO9hR13gd6ACsDmgOnb4AeJxEfkJ70fXKSDbKcB+EVkvIo8C5wKXq+qhaOyKiKjf1fDfX4TXJpuichjvD4yIrAfuwZO8/lREOgGrorXt219GugdLvu/lCOb7nIQ3fHK/yPH0is3w/hAEyRK8B3v/UtXUaI2p6i4RmQYMFJEjQEG8PwTfR2l3D/CGiLyV9kdARIYCpYGo/T6liXVXONwNyI8X2N7Gu81p5qCOmwjwNs23Kb7vvwC/ArUD9rkAcCWwFGgYoN2qQIt07/M4aG/Bu5VfBjQI2HZz4BFgVJDf50l1vAtUD9BeSeCvwNd4a8GjupXPpI609nbSJqfSlrBr5/0HP6oBPC0+yW4pvB/FLaoa1V//TOxfDszTgJMj+DMYugG/qOqKIG379k/o8QZtGzgH2KKqP7mowwUu28S3XwxvvH9vtoXDt30akE9VA7mrOJVJ2CDqEhEpqKqHHdl2+sMzDCNnsSBqGIYRBYn0dN4wDCPusCBqGIYRBRZEDcMwosCCqGEYRhRYEM0liEiqiCwWkR9F5L1oknmIyGsi0t9//ZKI1M+ibEc/oUi4dawV+bPmeGb7TyqzP8y67heRW8P10TBCwYJo7uGQqjZV1YZ4GZ2uTX9QRCJanaaqV6m3sigzOgJhB1HDyC1YEM2dzABO93uJM0RkPLBMRJJE5HERmSci36etgffXfT8jIitE5EvgeCJjEZkuIi391z1EZKGILBGRqSJSHS9Y3+T3gjuISDkR+cCvY56ItPfPLSMin4vIUhF5CW+VUpaIyMcissA/Z8RJx570908VkXL+vloiMtk/Z0Y0eQ8MI1QSZu28ERp+j7MnMNnf1RxvGegaPxDtUdVW4qXjmyUin+OtKa8D1Mdbd78MeOUku+WA/wFn+7ZKq+pOEXkB2K+q//bLvQk8qaozRaQa3rLFenjr2Geq6j9E5Dy8JarZcYVfRyFgnoh8oKo78NLQzVfVm0TkXt/29XjJQK5V1Z9FpA1e7s/OETSjYYSMBdHcQyERWey/noGfTBmYq6pr/P3nAo3TxjuBEng5Ws8G3lIvgcYmPwHGybQFvkmzpao7M/GjK1Bf/kivWVy8LPlnAxf6534mIrtC+Ex/FZEL/NdVfV93AMeAd/z944AP/TrOBN5LV3fgciyGcTIWRHMPh1S1afodfjA5kH4X8BdVnXJSuV4B+pEHaHvyslkJM2exeJn7uwLtVPWgiEzHy2iUEerXu/vkNjAM19iY6KnFFOD//GQliMgZ4mVM/wa42B8zrUTG6f9mA2eLSA3/3NL+/n14+S7T+Bz4S9obEUkLat/gJdFO04bKLsN8CWCXH0Dr4vWE08iDl8AY3+ZMP0nHGhEZ4NchIhJxnlnDCBULoqcWL+GNdy4UkR/xhOjy4smt/Owfex347uQTVXUbMALv1nkJf9xOfwpckPZgCS+FW0v/wdUy/pgl8ABeEF6Kd1v/aza+TgbyishyvBygs9MdO4AnV/Ij3pjnP/z9g4Erff+WAn1DaBPDiApLQGIYhhEF1hM1DMOIAguihmEYUWBB1DAMIwosiBqGYUSBBVHDMIwosCD6/xsFo2AUjAIKwGghOgpGwSgYBRQAAJHP0DNLblNDAAAAAElFTkSuQmCC\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "misclassified_idx=np.where(p_test!=y_test)[0]\n",
        "i=np.random.choice(misclassified_idx)\n",
        "plt.imshow(x_test[i],cmap='gray')\n",
        "plt.title('True label: %s  Predicted label: %s'%(y_test[i],p_test[i]))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 299
        },
        "id": "nRhU20Opq-Nc",
        "outputId": "0f5a231d-c9c8-4c7a-c7d0-fa241f5d3ebf"
      },
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'True label: 5  Predicted label: 6')"
            ]
          },
          "metadata": {},
          "execution_count": 21
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}