fashion_mnist.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "7iZoan3EPNJW"
      },
      "outputs": [],
      "source": [
        "import tensorflow as tf\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.metrics import classification_report"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "(xtrain,ytrain),(xtest,ytest) = tf.keras.datasets.fashion_mnist.load_data()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8ChGhaf7PXnD",
        "outputId": "24c2e4c2-d880-40f0-989f-9afc8cda949f",
        "collapsed": true
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-labels-idx1-ubyte.gz\n",
            "\u001b[1m29515/29515\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 0us/step\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-images-idx3-ubyte.gz\n",
            "\u001b[1m26421880/26421880\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 0us/step\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-labels-idx1-ubyte.gz\n",
            "\u001b[1m5148/5148\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 0us/step\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-images-idx3-ubyte.gz\n",
            "\u001b[1m4422102/4422102\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 0us/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.imshow(xtrain[0])\n",
        "plt.title(f\"object = {ytrain[0]}\")\n",
        "plt.colorbar()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 469
        },
        "id": "Sel8GqAsPob7",
        "outputId": "b7705b30-7ba6-48c8-db8e-6f56fce53101",
        "collapsed": true
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.colorbar.Colorbar at 0x7fcc46cf0280>"
            ]
          },
          "metadata": {},
          "execution_count": 4
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "xtrain[0].shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wVuALWHtPqgF",
        "outputId": "c8ab4a6a-f05a-4ffb-94c1-58cc2a95edbe",
        "collapsed": true
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(28, 28)"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "xtrain.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JUr-hO-kQDVZ",
        "outputId": "2e07b8f5-f5b3-4193-b8d8-f4a7cf4b96f4",
        "collapsed": true
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(60000, 28, 28)"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "label_to_remove = 8\n",
        "def remove_label(X, y, label):\n",
        "    mask = y != label  # Create a mask where the label is not equal to the target label\n",
        "    return X[mask], y[mask]\n",
        "\n",
        "# Remove class 8 from both training and testing datasets\n",
        "xtrain, ytrain = remove_label(xtrain, ytrain, label_to_remove)\n",
        "xtest, ytest = remove_label(xtest, ytest, label_to_remove)\n"
      ],
      "metadata": {
        "id": "xw1ZIYE-h1rE"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "label_to_remove = 6\n",
        "def remove_label(X, y, label):\n",
        "    mask = y != label  # Create a mask where the label is not equal to the target label\n",
        "    return X[mask], y[mask]\n",
        "\n",
        "# Remove class 8 from both training and testing datasets\n",
        "xtrain, ytrain = remove_label(xtrain, ytrain, label_to_remove)\n",
        "xtest, ytest = remove_label(xtest, ytest, label_to_remove)\n"
      ],
      "metadata": {
        "id": "8Sd431fAi7uW"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "class_images = {}\n",
        "for img, label in zip(xtrain, ytrain):\n",
        "    if label not in class_images:\n",
        "        class_images[label] = img\n",
        "    if len(class_images) == 10:  # Stop when we have all 10 classes\n",
        "        break"
      ],
      "metadata": {
        "id": "MocOH3vXQSUR"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "plt.figure(figsize=(10, 2))\n",
        "for i in range(10):\n",
        "    if i in class_images:\n",
        "      plt.imshow(class_images[i])\n",
        "      plt.title(f\"Class {i}\")\n",
        "      plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "collapsed": true,
        "id": "impez8UlQo4O",
        "outputId": "ffa02fa8-c45c-4681-965f-1ae50e182402"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x200 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAaAAAAGzCAYAAABpdMNsAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy81sbWrAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAkxklEQVR4nO3df3RU5YH/8c8kJEMCyYQQkslACAkiKD/SUwSKUMSSAmnXhUpdqf1uwdOF1Q2uSP1x2KOi7J7NLp5tbSlV2+NCe1a0ZbfiF23pQSxhUaCCUhalkcS0gCRBoknI7x/zfP9gnX5HfvlckjyT8H6dc88hM/cz95nLhU/uzJ1nfMYYIwAAelmc6wEAAK5OFBAAwAkKCADgBAUEAHCCAgIAOEEBAQCcoIAAAE5QQAAAJyggAIATFBBgadSoUVq6dKnrYQB9HgUE/K+Kigr97d/+rfLz8zVw4EClpqZqxowZ+v73v6+WlhbXw7ukxsZGrVmzRvPnz1d6erp8Pp82bdrkeljAJQ1wPQAgFrzyyiu67bbb5Pf79a1vfUsTJkxQe3u79uzZowceeEDvvPOOfvzjH7se5kWdOXNGa9eu1ciRI1VQUKBdu3a5HhJwWRQQrnqVlZVavHixcnNz9dprryk7OztyX3FxscrLy/XKK684HOHlZWdnq6qqSsFgUAcOHNCUKVNcDwm4LF6Cw1Vv3bp1amxs1LPPPhtVPp+45pprdO+99140/9FHH+n+++/XxIkTNXjwYKWmpqqoqEi///3vz1t3/fr1Gj9+vJKTkzVkyBDdcMMN2rx5c+T+s2fPauXKlRo1apT8fr8yMzP15S9/WW+99dYln4Pf71cwGLR41oB7nAHhqrdt2zbl5+frxhtv9JR///33tXXrVt12223Ky8tTTU2NnnnmGd1000169913FQqFJEk/+clP9Pd///f6+te/rnvvvVetra06fPiw9u/frzvuuEOSdNddd+k///M/tWLFCl1//fWqra3Vnj17dPToUX3+85/vtucMxAIKCFe1hoYGffDBB1qwYIHnx5g4caLee+89xcX9+QWFv/7rv9a4ceP07LPP6pFHHpF07n2m8ePHa8uWLRd9rFdeeUXLli3Tv/3bv0Vue/DBBz2PDYhlvASHq1pDQ4MkKSUlxfNj+P3+SPl0dXWptrZWgwcP1tixY6NeOktLS9PJkyf15ptvXvSx0tLStH//fp06dcrzeIC+ggLCVS01NVXSufdevAqHw/re976nMWPGyO/3KyMjQ8OGDdPhw4dVX18fWe+hhx7S4MGDNXXqVI0ZM0bFxcV6/fXXox5r3bp1OnLkiHJycjR16lQ99thjev/99z2PDYhlFBCuaqmpqQqFQjpy5Ijnx/jnf/5nrVq1SrNmzdJ//Md/6De/+Y127Nih8ePHKxwOR9a77rrrVFZWphdeeEEzZ87Uf/3Xf2nmzJlas2ZNZJ2/+qu/0vvvv6/169crFArpiSee0Pjx4/XrX//6ip4nEJMMcJVbvny5kWTeeOONz7R+bm6uWbJkSeTngoICc/PNN5+33vDhw81NN9100cdpa2szX/3qV018fLxpaWm54Do1NTVm+PDhZsaMGZ9pbMYY8+abbxpJZuPGjZ85A7jAGRCueg8++KAGDRqkv/mbv1FNTc1591dUVOj73//+RfPx8fEyxkTdtmXLFn3wwQdRt9XW1kb9nJiYqOuvv17GGHV0dKirqyvqJTtJyszMVCgUUltbm+3TAmIeV8Hhqjd69Ght3rxZt99+u6677rqomRDeeOMNbdmy5ZJzv/3FX/yF1q5dqzvvvFM33nij/ud//kfPPfec8vPzo9abO3eugsGgZsyYoaysLB09elQ//OEP9dWvflUpKSmqq6vTiBEj9PWvf10FBQUaPHiwXn31Vb355ptRV8VdzA9/+EPV1dVFLmDYtm2bTp48KUm65557FAgEvO8koCe4PgUDYsV7771nli1bZkaNGmUSExNNSkqKmTFjhlm/fr1pbW2NrPfpl+BaW1vNd77zHZOdnW2SkpLMjBkzzN69e81NN90U9RLcM888Y2bNmmWGDh1q/H6/GT16tHnggQdMfX29MebcS3IPPPCAKSgoMCkpKWbQoEGmoKDA/OhHP/pM48/NzTWSLrhUVlZ2xy4CupXPmE+9dgAAQC/gPSAAgBMUEADACQoIAOAEBQQAcIICAgA4QQEBAJyIuQ+ihsNhnTp1SikpKfL5fK6HAwCwZIzR2bNnFQqFor6m5NNiroBOnTqlnJwc18MAAFyhEydOaMSIERe9P+YK6JPvZZmpr2iAEhyPBt2t4+bPWWeOL7b/rLTv40TrjCQNaLQ/6/Z12mfa07usMzL220k+5e1V9ubxrdaZYen2X2lxzZAPrTOn53j/6gz0jk51aI9+ddnv2eqxAtqwYYOeeOIJVVdXq6CgQOvXr9fUqVMvm/vkZbcBStAAHwXU35gBA60zcUkeCqjFWwHFeSgTT5mk3imgeL+3AopLss/ED2q3ziQMsv974v+FPuB//8le7m2UHrkI4ec//7lWrVqlNWvW6K233lJBQYHmzZun06dP98TmAAB9UI8U0He/+10tW7ZMd955p66//no9/fTTSk5O1r//+7/3xOYAAH1QtxdQe3u7Dh48qMLCwj9vJC5OhYWF2rt373nrt7W1qaGhIWoBAPR/3V5AZ86cUVdXl7KysqJuz8rKUnV19Xnrl5SUKBAIRBaugAOAq4PzD6KuXr1a9fX1keXEiROuhwQA6AXdfhVcRkaG4uPjz/tq45qaGgWDwfPW9/v98vv93T0MAECM6/YzoMTERE2ePFk7d+6M3BYOh7Vz505Nnz69uzcHAOijeuRzQKtWrdKSJUt0ww03aOrUqXryySfV1NSkO++8syc2BwDog3qkgG6//XZ9+OGHevTRR1VdXa3Pfe5z2r59+3kXJgAArl49NhPCihUrtGLFip56ePRRH9xs/8n3mde+Y53pDMdbZyRp4bC3rDOjE+ynk5nst98Ph9vtp8f5Q7u3X/qOtgy3zrxzNts6s2DoIevMj5VvnUFscn4VHADg6kQBAQCcoIAAAE5QQAAAJyggAIATFBAAwAkKCADgBAUEAHCCAgIAOEEBAQCcoIAAAE5QQAAAJ3psMlLgQjqTjXXmdydyrTPZQxqsM5K04+MJ1pldcV3WmWetE1LagGbrTJzPfn9LUnnTMOvMH+vTrTNp2U3WmfjxY60zXe+UWWfQ8zgDAgA4QQEBAJyggAAATlBAAAAnKCAAgBMUEADACQoIAOAEBQQAcIICAgA4QQEBAJyggAAATlBAAAAnKCAAgBPMho1eFRj9sXXmmvQz1plQUr11RpKG++3HF0qos8682ZhnnfHHdVpnAh5m0JakjuR468wAX9g6kxLXap35oHCodSb4jnUEvYAzIACAExQQAMAJCggA4AQFBABwggICADhBAQEAnKCAAABOUEAAACcoIACAExQQAMAJCggA4AQFBABwgslI0avyh9RaZ3KS7CcIHe6vs85I0tiBp6wzv2/Otc54mVg0wddlnQkl2O87SQob+99N0wc0WWcGenhO7WnWEcQozoAAAE5QQAAAJyggAIATFBAAwAkKCADgBAUEAHCCAgIAOEEBAQCcoIAAAE5QQAAAJyggAIATFBAAwAkmI0WvyhtkPxnpgdqR1pn9XaOsM5L0rdxW68yEpBPWmQ87U60zXiR6mOzTq6r2gHUmXsY60zm22TqD2MQZEADACQoIAOBEtxfQY489Jp/PF7WMGzeuuzcDAOjjeuQ9oPHjx+vVV1/980YG8FYTACBajzTDgAEDFAwGe+KhAQD9RI+8B3Ts2DGFQiHl5+frm9/8po4fP37Rddva2tTQ0BC1AAD6v24voGnTpmnTpk3avn27nnrqKVVWVuqLX/yizp49e8H1S0pKFAgEIktOTk53DwkAEIO6vYCKiop02223adKkSZo3b55+9atfqa6uTr/4xS8uuP7q1atVX18fWU6csP9MBQCg7+nxqwPS0tJ07bXXqry8/IL3+/1++f3+nh4GACDG9PjngBobG1VRUaHs7Oye3hQAoA/p9gK6//77VVpaqj/+8Y9644039LWvfU3x8fH6xje+0d2bAgD0Yd3+EtzJkyf1jW98Q7W1tRo2bJhmzpypffv2adiwYd29KQBAH9btBfTCCy9090MiRsUlJ1tnRg20v8hk28cTrDOdnfHWGUn6qb5gnclJqbPOfCn9D9aZUQkfWmfK2kLWGcnbxKJH6+w/+/d68mjrzKgs+wltEZuYCw4A4AQFBABwggICADhBAQEAnKCAAABOUEAAACcoIACAExQQAMAJCggA4AQFBABwggICADhBAQEAnOjxL6RD/xUXzLTOHG9rts60fZRknfGnt1hnJGlwQrt1JjiwwTrTYewnS82Mb7TOPPzHqdYZSeoK+6wznV32z+lMR4p1Jk7GOoPYxBkQAMAJCggA4AQFBABwggICADhBAQEAnKCAAABOUEAAACcoIACAExQQAMAJCggA4AQFBABwggICADhBAQEAnGA2bHjWmZlqnTnbOdB+Q2H7SGJip31IUmNHonXGH2e/rf87PsM6M778A+vMjcFK64wk/fepfOtMS2uCdaayxX4/tHTab8d+PnX0Bs6AAABOUEAAACcoIACAExQQAMAJCggA4AQFBABwggICADhBAQEAnKCAAABOUEAAACcoIACAExQQAMAJJiOFZ11J9odPdUtKD4zkfP4BXZ5yWclnrTPv1Gfbb8hU2W+nbbh15qP2ZOuMJM3Itp/E9L2GTOtMS5f9xKLJCe3WGWOdQG/gDAgA4AQFBABwggICADhBAQEAnKCAAABOUEAAACcoIACAExQQAMAJCggA4AQFBABwggICADhBAQEAnGAyUnjnYYbHj9u8TY5pKz4u7Cl3qjFgnVmce8A682ulWWeONoWsMzXNqdYZSdp3Os86c32o2jqTltBinany8Jz4jy42cQYEAHCCAgIAOGFdQLt379Ytt9yiUCgkn8+nrVu3Rt1vjNGjjz6q7OxsJSUlqbCwUMeOHeuu8QIA+gnrAmpqalJBQYE2bNhwwfvXrVunH/zgB3r66ae1f/9+DRo0SPPmzVNra+sVDxYA0H9YvzdXVFSkoqKiC95njNGTTz6phx9+WAsWLJAk/exnP1NWVpa2bt2qxYsXX9loAQD9Rre+B1RZWanq6moVFhZGbgsEApo2bZr27t17wUxbW5saGhqiFgBA/9etBVRdfe4yzKysrKjbs7KyIvd9WklJiQKBQGTJycnpziEBAGKU86vgVq9erfr6+shy4sQJ10MCAPSCbi2gYDAoSaqpqYm6vaamJnLfp/n9fqWmpkYtAID+r1sLKC8vT8FgUDt37ozc1tDQoP3792v69OnduSkAQB9nfRVcY2OjysvLIz9XVlbq0KFDSk9P18iRI7Vy5Ur90z/9k8aMGaO8vDw98sgjCoVCWrhwYXeOGwDQx1kX0IEDB3TzzTdHfl61apUkacmSJdq0aZMefPBBNTU1afny5aqrq9PMmTO1fft2DRw4sPtGDQDo86wLaPbs2TLm4rNQ+nw+rV27VmvXrr2igaEP8NlHusK9c92L1+0MSmy3zoxKPONhS2nWidKTo60z/+eaN60zkvT0+7OtM2daBllnRg2utc50dMVbZ5iMNDY5vwoOAHB1ooAAAE5QQAAAJyggAIATFBAAwAkKCADgBAUEAHCCAgIAOEEBAQCcoIAAAE5QQAAAJyggAIATFBAAwAkmiYV3F58U/aLaPcxkHN9s/3tSU2uidUaSRg+xn9n6g44hnrZlq6U8YJ0Zcf1Hnrbla7Pf51Vn7MenYfaRhPgu+xBiEmdAAAAnKCAAgBMUEADACQoIAOAEBQQAcIICAgA4QQEBAJyggAAATlBAAAAnKCAAgBMUEADACQoIAOAEk5GiV4WNzzpjPByl7W0J9iFJg+LbrTNlzUEPW+qwTqT9wcNm/tJDRlJ8ept1xuezn532nbps64yXYwixiTMgAIATFBAAwAkKCADgBAUEAHCCAgIAOEEBAQCcoIAAAE5QQAAAJyggAIATFBAAwAkKCADgBAUEAHCCyUjhnYc5IZMT7Cfh/HBg2H5D7d5+t+ow9rlTLQEPWzpjncjc86F1JuGhTuuMJMXF2+9zE7Y/IFISWq0zH7UOss4gNnEGBABwggICADhBAQEAnKCAAABOUEAAACcoIACAExQQAMAJCggA4AQFBABwggICADhBAQEAnKCAAABOMBkpPAvH208+6fMZ+w3F22cGp7XYb0dSvIfxHTw2yjpzrYfJSHXmI/uMR8bDX5OXCUy9CBsPs+AiJnEGBABwggICADhhXUC7d+/WLbfcolAoJJ/Pp61bt0bdv3TpUvl8vqhl/vz53TVeAEA/YV1ATU1NKigo0IYNGy66zvz581VVVRVZnn/++SsaJACg/7G+CKGoqEhFRUWXXMfv9ysYDHoeFACg/+uR94B27dqlzMxMjR07Vnfffbdqa2svum5bW5saGhqiFgBA/9ftBTR//nz97Gc/086dO/Wv//qvKi0tVVFRkbq6ui64fklJiQKBQGTJycnp7iEBAGJQt38OaPHixZE/T5w4UZMmTdLo0aO1a9cuzZkz57z1V69erVWrVkV+bmhooIQA4CrQ45dh5+fnKyMjQ+Xl5Re83+/3KzU1NWoBAPR/PV5AJ0+eVG1trbKzs3t6UwCAPsT6JbjGxsaos5nKykodOnRI6enpSk9P1+OPP65FixYpGAyqoqJCDz74oK655hrNmzevWwcOAOjbrAvowIEDuvnmmyM/f/L+zZIlS/TUU0/p8OHD+ulPf6q6ujqFQiHNnTtX//iP/yi/3999owYA9HnWBTR79myZS8xU+Jvf/OaKBoS+I5xo/wpuur/ZOnPyI/trZRKDndYZSQok2E9imvhBgqdt2eqqtZ+MtDns7Re/AQPsJxaNi7OfwbS1y37ftXTYZ5KsE+gNzAUHAHCCAgIAOEEBAQCcoIAAAE5QQAAAJyggAIATFBAAwAkKCADgBAUEAHCCAgIAOEEBAQCcoIAAAE5QQAAAJ7r9K7mBS5kYOGWdOdo8xjqTlOBtNuzB8W3WmUCFp031isq2YZ5yfg/7r7k10TqTlmg/+/hHLcnWGcQmzoAAAE5QQAAAJyggAIATFBAAwAkKCADgBAUEAHCCAgIAOEEBAQCcoIAAAE5QQAAAJyggAIATFBAAwAkmI4Vnvi5jnclOqLPOtIY6rDNJCfYZSQoMaLbODHm30dO2esPRs0FPuezUBuvMsaZM60xYPvuMsc8gNnEGBABwggICADhBAQEAnKCAAABOUEAAACcoIACAExQQAMAJCggA4AQFBABwggICADhBAQEAnKCAAABOMBkpYl7coM5e21a87CdYHXC63jrTW8/oSHW2p1xR3rvWmcZ2v3VmUHy7daatg/+2+gvOgAAATlBAAAAnKCAAgBMUEADACQoIAOAEBQQAcIICAgA4QQEBAJyggAAATlBAAAAnKCAAgBMUEADACWb1Q69K8HVZZwYm2U9YOXRgk3VGkuq7kqwz4erTnrbVG1rOJHvKdeXZ/246MuVj68y1g2qsM2/HDbfOIDZxBgQAcIICAgA4YVVAJSUlmjJlilJSUpSZmamFCxeqrKwsap3W1lYVFxdr6NChGjx4sBYtWqSaGvvTbABA/2ZVQKWlpSouLta+ffu0Y8cOdXR0aO7cuWpq+vPr7ffdd5+2bdumLVu2qLS0VKdOndKtt97a7QMHAPRtVhchbN++PernTZs2KTMzUwcPHtSsWbNUX1+vZ599Vps3b9aXvvQlSdLGjRt13XXXad++ffrCF77QfSMHAPRpV/QeUH39ua8iTk9PlyQdPHhQHR0dKiwsjKwzbtw4jRw5Unv37r3gY7S1tamhoSFqAQD0f54LKBwOa+XKlZoxY4YmTJggSaqurlZiYqLS0tKi1s3KylJ1dfUFH6ekpESBQCCy5OTkeB0SAKAP8VxAxcXFOnLkiF544YUrGsDq1atVX18fWU6cOHFFjwcA6Bs8fRB1xYoVevnll7V7926NGDEicnswGFR7e7vq6uqizoJqamoUDAYv+Fh+v19+v9/LMAAAfZjVGZAxRitWrNCLL76o1157TXl5eVH3T548WQkJCdq5c2fktrKyMh0/flzTp0/vnhEDAPoFqzOg4uJibd68WS+99JJSUlIi7+sEAgElJSUpEAjo29/+tlatWqX09HSlpqbqnnvu0fTp07kCDgAQxaqAnnrqKUnS7Nmzo27fuHGjli5dKkn63ve+p7i4OC1atEhtbW2aN2+efvSjH3XLYAEA/YdVARljLrvOwIEDtWHDBm3YsMHzoNB/NYft3++Li7v8cfdpmf6z1hlJOt6Sbp0Jt3rbVm9I+CjeU+5sx0DrTGuX/VvKgfhm60xHl7fnhNjDXHAAACcoIACAExQQAMAJCggA4AQFBABwggICADhBAQEAnKCAAABOUEAAACcoIACAExQQAMAJCggA4AQFBABwwtM3ogKS1JVk//vLyfYh1hmfz3427GBig3VGkg6eybHODFbszoad8idvuUED2qwzH7cneduYpY4OZsPuLzgDAgA4QQEBAJyggAAATlBAAAAnKCAAgBMUEADACQoIAOAEBQQAcIICAgA4QQEBAJyggAAATlBAAAAnmIwUnnUl+KwzLV0J1pmBCZ3WmcCAZuuMJJ2pH2ydsU/0nuTTYU+5sLH/3dRLJiW+1X47YfvjDrGJMyAAgBMUEADACQoIAOAEBQQAcIICAgA4QQEBAJyggAAATlBAAAAnKCAAgBMUEADACQoIAOAEBQQAcILJSOFZ2NNkpIk9MJLzJfi6POU6Wnvpn4TPw4SaxlhHBrR4m4y0odNvnWkPx1tnyluzrDOdbfy31V9wBgQAcIICAgA4QQEBAJyggAAATlBAAAAnKCAAgBMUEADACQoIAOAEBQQAcIICAgA4QQEBAJyggAAATjCrH3rV+2eH9sp2Pmgf4ilnWnrnn4Qv3n7iTtPZaZ3x1zRbZySptSvBOhM29hOshmWfMa32+w6xiTMgAIATFBAAwAmrAiopKdGUKVOUkpKizMxMLVy4UGVlZVHrzJ49Wz6fL2q56667unXQAIC+z6qASktLVVxcrH379mnHjh3q6OjQ3Llz1dTUFLXesmXLVFVVFVnWrVvXrYMGAPR9Vu+4bt++PernTZs2KTMzUwcPHtSsWbMitycnJysYDHbPCAEA/dIVvQdUX18vSUpPT4+6/bnnnlNGRoYmTJig1atXq7n54lfitLW1qaGhIWoBAPR/nq85DYfDWrlypWbMmKEJEyZEbr/jjjuUm5urUCikw4cP66GHHlJZWZl++ctfXvBxSkpK9Pjjj3sdBgCgj/JcQMXFxTpy5Ij27NkTdfvy5csjf544caKys7M1Z84cVVRUaPTo0ec9zurVq7Vq1arIzw0NDcrJyfE6LABAH+GpgFasWKGXX35Zu3fv1ogRIy657rRp0yRJ5eXlFywgv98vv9/vZRgAgD7MqoCMMbrnnnv04osvateuXcrLy7ts5tChQ5Kk7OxsTwMEAPRPVgVUXFyszZs366WXXlJKSoqqq6slSYFAQElJSaqoqNDmzZv1la98RUOHDtXhw4d13333adasWZo0aVKPPAEAQN9kVUBPPfWUpHMfNv3/bdy4UUuXLlViYqJeffVVPfnkk2pqalJOTo4WLVqkhx9+uNsGDADoH6xfgruUnJwclZaWXtGAAABXB2bDhmfNw+w/RjYlrco6U1afZZ3JGNBonZEkX1svTY/oYTZseZgNO67dPiNJgYRW64yX2bAD8S3WmfjB3p4TYg+TkQIAnKCAAABOUEAAACcoIACAExQQAMAJCggA4AQFBABwggICADhBAQEAnKCAAABOUEAAACcoIACAE0xGCs+G/d5+wspfh26wzpgBl56F/UKeyQtYZyRpeKn9tjzp6uqVzZg/nvSU++8/5VtnMgP2E8AeiMu1ziS+m2SdQWziDAgA4AQFBABwggICADhBAQEAnKCAAABOUEAAACcoIACAExQQAMAJCggA4AQFBABwggICADgRc3PBGXNuLq5OdUi9NC0XvOnqtJ8LLtxq/5fqZS64ruY264wkdXbY/5PoNB3WGZ/xsB9Mp3UmzrRbZySpq9n+77ZzgP0+7/DZj6+rzcPYPPwdwbtOndvf5jLHuc9cbo1edvLkSeXk5LgeBgDgCp04cUIjRoy46P0xV0DhcFinTp1SSkqKfD5f1H0NDQ3KycnRiRMnlJqa6miE7rEfzmE/nMN+OIf9cE4s7AdjjM6ePatQKKS4uIu/0xNzL8HFxcVdsjElKTU19ao+wD7BfjiH/XAO++Ec9sM5rvdDIHD5r0ThIgQAgBMUEADAiT5VQH6/X2vWrJHf73c9FKfYD+ewH85hP5zDfjinL+2HmLsIAQBwdehTZ0AAgP6DAgIAOEEBAQCcoIAAAE5QQAAAJ/pMAW3YsEGjRo3SwIEDNW3aNP3ud79zPaRe99hjj8nn80Ut48aNcz2sHrd7927dcsstCoVC8vl82rp1a9T9xhg9+uijys7OVlJSkgoLC3Xs2DE3g+1Bl9sPS5cuPe/4mD9/vpvB9pCSkhJNmTJFKSkpyszM1MKFC1VWVha1Tmtrq4qLizV06FANHjxYixYtUk1NjaMR94zPsh9mz5593vFw1113ORrxhfWJAvr5z3+uVatWac2aNXrrrbdUUFCgefPm6fTp066H1uvGjx+vqqqqyLJnzx7XQ+pxTU1NKigo0IYNGy54/7p16/SDH/xATz/9tPbv369BgwZp3rx5am21nzU5ll1uP0jS/Pnzo46P559/vhdH2PNKS0tVXFysffv2aceOHero6NDcuXPV1NQUWee+++7Ttm3btGXLFpWWlurUqVO69dZbHY66+32W/SBJy5Ytizoe1q1b52jEF2H6gKlTp5ri4uLIz11dXSYUCpmSkhKHo+p9a9asMQUFBa6H4ZQk8+KLL0Z+DofDJhgMmieeeCJyW11dnfH7/eb55593MMLe8en9YIwxS5YsMQsWLHAyHldOnz5tJJnS0lJjzLm/+4SEBLNly5bIOkePHjWSzN69e10Ns8d9ej8YY8xNN91k7r33XneD+gxi/gyovb1dBw8eVGFhYeS2uLg4FRYWau/evQ5H5saxY8cUCoWUn5+vb37zmzp+/LjrITlVWVmp6urqqOMjEAho2rRpV+XxsWvXLmVmZmrs2LG6++67VVtb63pIPaq+vl6SlJ6eLkk6ePCgOjo6oo6HcePGaeTIkf36ePj0fvjEc889p4yMDE2YMEGrV69Wc3Ozi+FdVMzNhv1pZ86cUVdXl7KysqJuz8rK0h/+8AdHo3Jj2rRp2rRpk8aOHauqqio9/vjj+uIXv6gjR44oJSXF9fCcqK6ulqQLHh+f3He1mD9/vm699Vbl5eWpoqJC//AP/6CioiLt3btX8fHxrofX7cLhsFauXKkZM2ZowoQJks4dD4mJiUpLS4tatz8fDxfaD5J0xx13KDc3V6FQSIcPH9ZDDz2ksrIy/fKXv3Q42mgxX0D4s6KiosifJ02apGnTpik3N1e/+MUv9O1vf9vhyBALFi9eHPnzxIkTNWnSJI0ePVq7du3SnDlzHI6sZxQXF+vIkSNXxfugl3Kx/bB8+fLInydOnKjs7GzNmTNHFRUVGj16dG8P84Ji/iW4jIwMxcfHn3cVS01NjYLBoKNRxYa0tDRde+21Ki8vdz0UZz45Bjg+zpefn6+MjIx+eXysWLFCL7/8sn77299GfX9YMBhUe3u76urqotbvr8fDxfbDhUybNk2SYup4iPkCSkxM1OTJk7Vz587IbeFwWDt37tT06dMdjsy9xsZGVVRUKDs72/VQnMnLy1MwGIw6PhoaGrR///6r/vg4efKkamtr+9XxYYzRihUr9OKLL+q1115TXl5e1P2TJ09WQkJC1PFQVlam48eP96vj4XL74UIOHTokSbF1PLi+CuKzeOGFF4zf7zebNm0y7777rlm+fLlJS0sz1dXVrofWq77zne+YXbt2mcrKSvP666+bwsJCk5GRYU6fPu16aD3q7Nmz5u233zZvv/22kWS++93vmrffftv86U9/MsYY8y//8i8mLS3NvPTSS+bw4cNmwYIFJi8vz7S0tDgeefe61H44e/asuf/++83evXtNZWWlefXVV83nP/95M2bMGNPa2up66N3m7rvvNoFAwOzatctUVVVFlubm5sg6d911lxk5cqR57bXXzIEDB8z06dPN9OnTHY66+11uP5SXl5u1a9eaAwcOmMrKSvPSSy+Z/Px8M2vWLMcjj9YnCsgYY9avX29GjhxpEhMTzdSpU82+fftcD6nX3X777SY7O9skJiaa4cOHm9tvv92Ul5e7HlaP++1vf2sknbcsWbLEGHPuUuxHHnnEZGVlGb/fb+bMmWPKysrcDroHXGo/NDc3m7lz55phw4aZhIQEk5uba5YtW9bvfkm70POXZDZu3BhZp6Wlxfzd3/2dGTJkiElOTjZf+9rXTFVVlbtB94DL7Yfjx4+bWbNmmfT0dOP3+80111xjHnjgAVNfX+924J/C9wEBAJyI+feAAAD9EwUEAHCCAgIAOEEBAQCcoIAAAE5QQAAAJyggAIATFBAAwAkKCADgBAUEAHCCAgIAOPH/AEZTfKS0Ha8mAAAAAElFTkSuQmCC\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "NORMALIZATION"
      ],
      "metadata": {
        "id": "ygQOAxTIWemr"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "xtrain,xtest = xtrain/255,xtest/255"
      ],
      "metadata": {
        "id": "DPHha_vrWFEn"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "NEURAL NETWORKS USING TENSORFLOW ANS KERAS."
      ],
      "metadata": {
        "id": "H_KbXXgqWhsM"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "model = tf.keras.Sequential([\n",
        "    tf.keras.Input(shape = (28,28)),\n",
        "    tf.keras.layers.Flatten(),\n",
        "    tf.keras.layers.Dense(128,activation = 'relu'),\n",
        "    tf.keras.layers.Dense(10,activation = 'softmax')\n",
        "])\n",
        "model.compile(optimizer = 'adam', metrics = ['accuracy'],loss = \"sparse_categorical_crossentropy\")\n",
        "model.fit(xtrain,ytrain,epochs = 20,verbose = 1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JoFAjVZMSlfk",
        "outputId": "6ab2dca0-07eb-49f7-e618-8b296f184c41",
        "collapsed": true
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/20\n",
            "\u001b[1m1500/1500\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 3ms/step - accuracy: 0.8401 - loss: 0.4638\n",
            "Epoch 2/20\n",
            "\u001b[1m1500/1500\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 3ms/step - accuracy: 0.9057 - loss: 0.2652\n",
            "Epoch 3/20\n",
            "\u001b[1m1500/1500\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 4ms/step - accuracy: 0.9188 - loss: 0.2262\n",
            "Epoch 4/20\n",
            "\u001b[1m1500/1500\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 3ms/step - accuracy: 0.9272 - loss: 0.2055\n",
            "Epoch 5/20\n",
            "\u001b[1m1500/1500\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 4ms/step - accuracy: 0.9299 - loss: 0.1968\n",
            "Epoch 6/20\n",
            "\u001b[1m1500/1500\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 4ms/step - accuracy: 0.9378 - loss: 0.1730\n",
            "Epoch 7/20\n",
            "\u001b[1m1500/1500\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 3ms/step - accuracy: 0.9379 - loss: 0.1684\n",
            "Epoch 8/20\n",
            "\u001b[1m1500/1500\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 5ms/step - accuracy: 0.9402 - loss: 0.1620\n",
            "Epoch 9/20\n",
            "\u001b[1m1500/1500\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 3ms/step - accuracy: 0.9431 - loss: 0.1527\n",
            "Epoch 10/20\n",
            "\u001b[1m1500/1500\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 3ms/step - accuracy: 0.9460 - loss: 0.1449\n",
            "Epoch 11/20\n",
            "\u001b[1m1500/1500\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 4ms/step - accuracy: 0.9484 - loss: 0.1395\n",
            "Epoch 12/20\n",
            "\u001b[1m1500/1500\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 3ms/step - accuracy: 0.9506 - loss: 0.1342\n",
            "Epoch 13/20\n",
            "\u001b[1m1500/1500\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 3ms/step - accuracy: 0.9517 - loss: 0.1274\n",
            "Epoch 14/20\n",
            "\u001b[1m1500/1500\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 4ms/step - accuracy: 0.9543 - loss: 0.1224\n",
            "Epoch 15/20\n",
            "\u001b[1m1500/1500\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m9s\u001b[0m 3ms/step - accuracy: 0.9566 - loss: 0.1175\n",
            "Epoch 16/20\n",
            "\u001b[1m1500/1500\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 3ms/step - accuracy: 0.9561 - loss: 0.1147\n",
            "Epoch 17/20\n",
            "\u001b[1m1500/1500\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 4ms/step - accuracy: 0.9590 - loss: 0.1106\n",
            "Epoch 18/20\n",
            "\u001b[1m1500/1500\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 3ms/step - accuracy: 0.9611 - loss: 0.1057\n",
            "Epoch 19/20\n",
            "\u001b[1m1500/1500\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 3ms/step - accuracy: 0.9602 - loss: 0.1050\n",
            "Epoch 20/20\n",
            "\u001b[1m1500/1500\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 4ms/step - accuracy: 0.9620 - loss: 0.1018\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.src.callbacks.history.History at 0x7c9c3cc37730>"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "loss , accuracy = model.evaluate(xtest,ytest)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8c7bu4w7UOtW",
        "outputId": "a012ec9a-e8ce-448d-c286-705eaac7f83d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[1m250/250\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9290 - loss: 0.2425\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "CLASSIFICATION REPORT ON TEST DATA"
      ],
      "metadata": {
        "id": "2HPOjQYolKsV"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "y_pred = model.predict(xtest)\n",
        "y_pred_classes = y_pred.argmax(axis=1)\n",
        "print(classification_report(ytest, y_pred_classes))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "collapsed": true,
        "id": "CCqt9hj6k_gJ",
        "outputId": "7b1e9e24-d621-4b80-80b3-9c4d3b6e5e0f"
      },
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[1m250/250\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.96      0.92      0.94      1000\n",
            "           1       0.99      0.98      0.98      1000\n",
            "           2       0.81      0.89      0.85      1000\n",
            "           3       0.93      0.91      0.92      1000\n",
            "           4       0.86      0.84      0.85      1000\n",
            "           5       0.98      0.97      0.98      1000\n",
            "           7       0.95      0.97      0.96      1000\n",
            "           9       0.96      0.96      0.96      1000\n",
            "\n",
            "    accuracy                           0.93      8000\n",
            "   macro avg       0.93      0.93      0.93      8000\n",
            "weighted avg       0.93      0.93      0.93      8000\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "np.argmax(model.predict(xtest)[0])"
      ],
      "metadata": {
        "id": "1yxqfJzrXwoH",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "4682f3b6-0413-4236-9d50-cf73773efd9f",
        "collapsed": true
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[1m250/250\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "9"
            ]
          },
          "metadata": {},
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "TEST WITH ANY NEW IMAGE"
      ],
      "metadata": {
        "id": "-R43ds1-DkCT"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "test_image = tf.keras.preprocessing.image.load_img(\"NEW_IMAGE_PATH\")\n",
        "test_image = tf.image.rgb_to_grayscale(test_image)\n",
        "test_image = tf.keras.preprocessing.image.img_to_array(test_image)\n",
        "test_image = tf.image.resize(test_image,[28,28])\n",
        "test_image = test_image.numpy().squeeze()\n",
        "test_image = test_image.astype(\"float64\")\n",
        "plt.imshow(test_image)\n",
        "test_image = test_image.reshape(1,28,28)\n",
        "test_image = test_image/255\n",
        "plt.title(f\"ID = {np.argmax(model.predict(test_image))}\")"
      ],
      "metadata": {
        "id": "slM1hBFJ_3g8"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}