{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "id": "oNzzFpapBz5f"
   },
   "outputs": [],
   "source": [
    "from keras.datasets import fashion_mnist\n",
    "from sklearn.metrics import classification_report,accuracy_score  \n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import recall_score, confusion_matrix,f1_score\n",
    "import numpy as np\n",
    "from sklearn.metrics import roc_curve, auc\n",
    "import matplotlib.pyplot as plt\n",
    "from collections import Counter\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "import itertools"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "id": "SqLSdbhnChN8"
   },
   "outputs": [],
   "source": [
    "(x_train,y_train),(x_test,y_test) = fashion_mnist.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "jAfjqrQZCqwG",
    "outputId": "a3345160-54f8-489f-a05c-6315bf36cd64"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(60000, 28, 28)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Ndm93jbkC1GM",
    "outputId": "cf1634cf-d1ce-4770-dfd9-21e88f8ad8bc"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10000, 28, 28)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "ULaKNvdQC3J6",
    "outputId": "5a9fa7bb-e06c-428e-fa9e-02a4daa79186"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Counter({9: 6000,\n",
       "         0: 6000,\n",
       "         3: 6000,\n",
       "         2: 6000,\n",
       "         7: 6000,\n",
       "         5: 6000,\n",
       "         1: 6000,\n",
       "         6: 6000,\n",
       "         4: 6000,\n",
       "         8: 6000})"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Counter(y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# EDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 296
    },
    "id": "EqRkwVlrDFMF",
    "outputId": "e1635f69-a188-456a-9d64-caf41bc22c58"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'class count')"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.bar(Counter(y_train).keys(),Counter(y_train).values())\n",
    "plt.xlabel('class name')\n",
    "plt.ylabel('class count')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 282
    },
    "id": "CKiT7c8MDPx8",
    "outputId": "e3426293-05ff-451f-91d8-c954a2ea639e"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<BarContainer object of 4 artists>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD4CAYAAAAXUaZHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAN70lEQVR4nO3dXYxc5X3H8e+PlyQqobzEGwuM20WR08oRikErgkrUUhE1QKQYlBaBVGIQknMBVajSC5IbkAgtaUuQEAmqERaOSpPQkgSroKTUAhFoAlkI4TWABabYNbBpKkKKSoX592KPy2DWnt2dnR3vw/cjjfbMM+fMPHNkf318PHOcqkKS1JYDRj0BSdLCM+6S1CDjLkkNMu6S1CDjLkkNOmjUEwBYtmxZjY+Pj3oakrSkPPjgg7+oqrGZHtsv4j4+Ps7k5OSopyFJS0qS5/f2mKdlJKlBxl2SGmTcJalBxl2SGmTcJalBxl2SGmTcJalBxl2SGmTcJalB+8U3VDVa45fePuopjNS2qz416ilIC67vkXuSlUnuSvJEkseTfL4bvzzJjiQPd7czerb5YpKtSZ5K8slhvgFJ0jvN5sj9DeALVfVQkkOBB5Pc2T12TVX9be/KSVYD5wAfAY4G/jXJh6tq10JOXJK0d32P3KtqZ1U91C2/CjwJrNjHJmuBb1XV61X1HLAVOHEhJitJmp05/YNqknHgeOD+bujiJI8k2ZjkiG5sBfBCz2bbmeEPgyTrk0wmmZyamprzxCVJezfruCd5P3ArcElV/Qq4HvgQsAbYCVw9lxeuqg1VNVFVE2NjM16OWJI0T7OKe5KDmQ77zVX1HYCqeqmqdlXVm8ANvHXqZQewsmfzY7oxSdIimc2nZQLcCDxZVV/tGT+qZ7WzgMe65c3AOUnem+RYYBXwwMJNWZLUz2w+LXMycB7waJKHu7EvAecmWQMUsA34HEBVPZ7kFuAJpj9pc5GflJGkxdU37lV1L5AZHrpjH9tcCVw5wLwkSQPw8gOS1CDjLkkNMu6S1CDjLkkNMu6S1CDjLkkNMu6S1CDjLkkNMu6S1CDjLkkNMu6S1CDjLkkNMu6S1CDjLkkNMu6S1CDjLkkNMu6S1CDjLkkNMu6S1CDjLkkNMu6S1CDjLkkNMu6S1CDjLkkNMu6S1CDjLkkNMu6S1CDjLkkNMu6S1CDjLkkNMu6S1KC+cU+yMsldSZ5I8niSz3fjRya5M8kz3c8juvEkuTbJ1iSPJDlh2G9CkvR2szlyfwP4QlWtBk4CLkqyGrgU2FJVq4At3X2A04FV3W09cP2Cz1qStE99415VO6vqoW75VeBJYAWwFtjUrbYJOLNbXgt8o6b9GDg8yVELPnNJ0l7N6Zx7knHgeOB+YHlV7eweehFY3i2vAF7o2Wx7N7bnc61PMplkcmpqao7TliTty6zjnuT9wK3AJVX1q97HqqqAmssLV9WGqpqoqomxsbG5bCpJ6mNWcU9yMNNhv7mqvtMNv7T7dEv38+VufAewsmfzY7oxSdIimc2nZQLcCDxZVV/teWgzsK5bXgfc1jP+2e5TMycBr/ScvpEkLYKDZrHOycB5wKNJHu7GvgRcBdyS5ELgeeDs7rE7gDOArcBrwAULOmNJUl99415V9wLZy8OnzrB+ARcNOC9J0gD8hqokNci4S1KDjLskNci4S1KDjLskNci4S1KDjLskNci4S1KDjLskNci4S1KDjLskNci4S1KDjLskNci4S1KDjLskNci4S1KDjLskNci4S1KDjLskNci4S1KDjLskNci4S1KDjLskNci4S1KDjLskNci4S1KDjLskNci4S1KDjLskNci4S1KDjLskNci4S1KD+sY9ycYkLyd5rGfs8iQ7kjzc3c7oeeyLSbYmeSrJJ4c1cUnS3s3myP0m4LQZxq+pqjXd7Q6AJKuBc4CPdNt8PcmBCzVZSdLs9I17Vd0D/HKWz7cW+FZVvV5VzwFbgRMHmJ8kaR4GOed+cZJHutM2R3RjK4AXetbZ3o29Q5L1SSaTTE5NTQ0wDUnSnuYb9+uBDwFrgJ3A1XN9gqraUFUTVTUxNjY2z2lIkmYyr7hX1UtVtauq3gRu4K1TLzuAlT2rHtONSZIW0bzinuSonrtnAbs/SbMZOCfJe5McC6wCHhhsipKkuTqo3wpJvgmcAixLsh24DDglyRqggG3A5wCq6vEktwBPAG8AF1XVruFMXZK0N33jXlXnzjB84z7WvxK4cpBJSZIG4zdUJalBxl2SGmTcJalBxl2SGmTcJalBxl2SGmTcJalBxl2SGmTcJalBxl2SGmTcJalBxl2SGmTcJalBxl2SGmTcJalBxl2SGmTcJalBxl2SGmTcJalBxl2SGmTcJalBxl2SGmTcJalBxl2SGmTcJalBxl2SGmTcJalBxl2SGmTcJalBxl2SGmTcJalBfeOeZGOSl5M81jN2ZJI7kzzT/TyiG0+Sa5NsTfJIkhOGOXlJ0sxmc+R+E3DaHmOXAluqahWwpbsPcDqwqrutB65fmGlKkuaib9yr6h7gl3sMrwU2dcubgDN7xr9R034MHJ7kqIWarCRpduZ7zn15Ve3sll8ElnfLK4AXetbb3o1JkhbRwP+gWlUF1Fy3S7I+yWSSyampqUGnIUnqMd+4v7T7dEv38+VufAewsme9Y7qxd6iqDVU1UVUTY2Nj85yGJGkm8437ZmBdt7wOuK1n/LPdp2ZOAl7pOX0jSVokB/VbIck3gVOAZUm2A5cBVwG3JLkQeB44u1v9DuAMYCvwGnDBEOYsSeqjb9yr6ty9PHTqDOsWcNGgk5IkDcZvqEpSg4y7JDXIuEtSg/qec5ekYRq/9PZRT2Gktl31qaE8r0fuktQg4y5JDTLuktQg4y5JDTLuktQg4y5JDTLuktQg4y5JDTLuktQg4y5JDTLuktQg4y5JDTLuktSgJX9VSK8oN5wrykla2jxyl6QGGXdJapBxl6QGGXdJapBxl6QGGXdJapBxl6QGGXdJapBxl6QGGXdJapBxl6QGGXdJapBxl6QGGXdJatBAl/xNsg14FdgFvFFVE0mOBL4NjAPbgLOr6r8Gm6YkaS4W4sj9D6tqTVVNdPcvBbZU1SpgS3dfkrSIhnFaZi2wqVveBJw5hNeQJO3DoHEv4F+SPJhkfTe2vKp2dssvAstn2jDJ+iSTSSanpqYGnIYkqdeg/83ex6tqR5IPAncm+Xnvg1VVSWqmDatqA7ABYGJiYsZ1JEnzM9CRe1Xt6H6+DHwXOBF4KclRAN3PlwedpCRpbuYd9ySHJDl09zLwR8BjwGZgXbfaOuC2QScpSZqbQU7LLAe+m2T38/xDVX0/yU+AW5JcCDwPnD34NCVJczHvuFfVs8BHZxj/T+DUQSYlSRqM31CVpAYZd0lqkHGXpAYZd0lqkHGXpAYZd0lqkHGXpAYZd0lqkHGXpAYZd0lqkHGXpAYZd0lqkHGXpAYZd0lqkHGXpAYZd0lqkHGXpAYZd0lqkHGXpAYZd0lqkHGXpAYZd0lqkHGXpAYZd0lqkHGXpAYZd0lqkHGXpAYZd0lqkHGXpAYZd0lqkHGXpAYZd0lq0NDinuS0JE8l2Zrk0mG9jiTpnYYS9yQHAl8DTgdWA+cmWT2M15IkvdOwjtxPBLZW1bNV9b/At4C1Q3otSdIeDhrS864AXui5vx34WO8KSdYD67u7v07y1JDmMmzLgF+M6sXzlVG98oJyHw5mpPuvAUv5199v7+2BYcW9r6raAGwY1esvlCSTVTUx6nksZe7Dwbj/BtPq/hvWaZkdwMqe+8d0Y5KkRTCsuP8EWJXk2CTvAc4BNg/ptSRJexjKaZmqeiPJxcAPgAOBjVX1+DBeaz+w5E8t7Qfch4Nx/w2myf2Xqhr1HCRJC8xvqEpSg4y7JDXIuEtLXJJPe4mPuUtySZLf2Mtj5ye5brHntJCMu7TEVdXmqrpq1PNYgi4BZox7C4x7jyTjSX6e5KYkTye5OcknktyX5JkkJyY5JMnGJA8k+WmStT3b/jDJQ93t97rxU5LcneSfuue+OUlG+04Xxyz354lJftTty39L8jvdtn+eZGO3fFySx/Z2lNWyWe7D/z/K7Na7ttuXzyb541G/h/1B9/v29iQ/634tXQYcDdyV5K5unQu6ffwAcPJIJ7wQqspbdwPGgTeA45j+g+9BYCMQpq+N8z3gL4E/7dY/HHgaOITpI4D3deOrgMlu+RTgFaa/yHUA8CPg46N+r/vR/vxN4KBu/U8At3bLBwD3AGcBk8DJo34/+/E+PB+4rlv/JuAfu3VXM32Np5G/j1HfgM8AN/TcPwzYBizr7h8F/DswBrwHuG/3Pl2qt5FdfmA/9lxVPQqQ5HFgS1VVkkeZ/o12DPDpJH/Rrf8+4LeA/wCuS7IG2AV8uOc5H6iq7d1zPtw9z72L8F72B/3252HApiSrgAIOBqiqN5OcDzwC/F1V3TeKye8n+u3DPX2vqt4EnkiyfBHnuT97FLg6yVeAf66qH+7xF+iPAXdX1RRAkm/z9t/DS45xf6fXe5bf7Ln/JtP7axfwmap624XOklwOvAR8lOmjpv/Zy3Pu4t213/vtzyuAu6rqrCTjwN09668Cfs30X5/fzfrtw32t/644BdhPVT2d5ATgDODLSbaMek7D5jn3ufsB8Ge7z5snOb4bPwzY2R0xncf0N3PV32G8dd2h83cPJjkMuBb4feADnjvWIJIcDbxWVX8P/A1wAvAqcGi3yv3AHyT5QJKDgT8ZzUwXjnGfuyuYPnXwSPdX5Cu68a8D65L8DPhd4L9HNL+l5q+Bv0ryU95+FHoN8LWqehq4ELgqyQdHMUE14Tjgge606GXAl5m+7MD3k9xVVTuBy5n+N7H7gCdHNdGF4uUHJKlBHrlLUoOMuyQ1yLhLUoOMuyQ1yLhLUoOMuyQ1yLhLUoP+D33f2Dv+z5joAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "mean_value=np.mean(x_train)\n",
    "max_value=np.max(x_train)\n",
    "min_value=np.min(x_train)\n",
    "std_value=np.std(x_train)\n",
    "plt.bar(['mean','max','min','std'],[mean_value,max_value,min_value,std_value])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 491
    },
    "id": "p9Wcxz-CDnLr",
    "outputId": "987d9d92-359a-41d8-8c94-f2f890de89d8"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-14-6e77c19ecad5>:5: MatplotlibDeprecationWarning: Adding an axes using the same arguments as a previous axes currently reuses the earlier instance.  In a future version, a new instance will always be created and returned.  Meanwhile, this warning can be suppressed, and the future behavior ensured, by passing a unique label to each axes instance.\n",
      "  plt.subplot(2,5,count+1)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2160x720 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(30, 10))\n",
    "count=0\n",
    "for i,j in zip(x_train,y_train):\n",
    "    if count<=9:\n",
    "      plt.subplot(2,5,count+1)\n",
    "      if j==count:\n",
    "        plt.imshow(i, cmap=plt.cm.gray_r, interpolation='nearest',label='en')\n",
    "        plt.text(10,35,j, fontsize=20)\n",
    "        count+=1\n",
    "    else:\n",
    "      break\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "id": "Ycq8irQAEERo"
   },
   "outputs": [],
   "source": [
    "s=StandardScaler()\n",
    "s.fit(x_train.reshape(-1,28*28))\n",
    "x_train_scale=s.transform(x_train.reshape(-1,28*28))\n",
    "x_test_scale=s.transform(x_test.reshape(-1,28*28))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 282
    },
    "id": "est9XpZrEcW4",
    "outputId": "84144f62-cdf9-4582-a0fd-0d38ce71a6c0"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<BarContainer object of 4 artists>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "mean_value=np.mean(x_train_scale)\n",
    "max_value=np.max(x_train_scale)\n",
    "min_value=np.min(x_train_scale)\n",
    "std_value=np.std(x_train_scale)\n",
    "plt.bar(['mean','max','min','std'],[mean_value,max_value,min_value,std_value])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "id": "SG2Laet0I6Hf"
   },
   "outputs": [],
   "source": [
    "def plot_confusion_matrix(cm, classes,\n",
    "                          title='Confusion matrix',  \n",
    "                          cmap=plt.cm.Blues):\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=0)\n",
    "    plt.yticks(tick_marks, classes)\n",
    "\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, cm[i, j],\n",
    "                 horizontalalignment=\"center\",\n",
    "                 color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "\n",
    "    plt.tight_layout()\n",
    "    plt.ylabel('True label')\n",
    "    plt.xlabel('Predicted label')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# KNN model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "75SF4Ja4EnRm",
    "outputId": "49ce96ee-b5c5-4b86-d3a5-1b5e9316b3a2"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'n_neighbors': 5}\n"
     ]
    }
   ],
   "source": [
    "parms_grid={\n",
    "    'n_neighbors':[1,5,10,20]\n",
    "}\n",
    "knn =KNeighborsClassifier()\n",
    "model = GridSearchCV(knn,parms_grid, cv=5)\n",
    "model.fit(x_train_scale, y_train)\n",
    "\n",
    "print(model.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Os2PsOPJILGQ",
    "outputId": "b17e736f-c6d1-4396-8b36-5b72b3ed24df"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'mean_fit_time': array([0.1545094 , 0.1340672 , 0.15557289, 0.1596209 ]),\n",
       " 'std_fit_time': array([0.02613362, 0.01451174, 0.03012524, 0.03136509]),\n",
       " 'mean_score_time': array([33.3480464 , 36.90351281, 37.3830595 , 37.73826046]),\n",
       " 'std_score_time': array([1.8121195 , 0.55499432, 1.4912456 , 1.52265049]),\n",
       " 'param_n_neighbors': masked_array(data=[1, 5, 10, 20],\n",
       "              mask=[False, False, False, False],\n",
       "        fill_value='?',\n",
       "             dtype=object),\n",
       " 'params': [{'n_neighbors': 1},\n",
       "  {'n_neighbors': 5},\n",
       "  {'n_neighbors': 10},\n",
       "  {'n_neighbors': 20}],\n",
       " 'split0_test_score': array([0.84675   , 0.85658333, 0.85433333, 0.84733333]),\n",
       " 'split1_test_score': array([0.84858333, 0.84975   , 0.84991667, 0.84108333]),\n",
       " 'split2_test_score': array([0.84733333, 0.85558333, 0.85541667, 0.84658333]),\n",
       " 'split3_test_score': array([0.85183333, 0.85733333, 0.85591667, 0.84891667]),\n",
       " 'split4_test_score': array([0.846     , 0.85341667, 0.85016667, 0.84208333]),\n",
       " 'mean_test_score': array([0.8481    , 0.85453333, 0.85315   , 0.8452    ]),\n",
       " 'std_test_score': array([0.00204898, 0.00273028, 0.00259026, 0.00306395]),\n",
       " 'rank_test_score': array([3, 1, 2, 4], dtype=int32)}"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.cv_results_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "23_G1zO2IOmP",
    "outputId": "a8b0c069-9fda-42a7-c8a4-5c7d5eb83bc3"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "class report：\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.77      0.86      0.81      1000\n",
      "           1       0.99      0.96      0.98      1000\n",
      "           2       0.74      0.80      0.77      1000\n",
      "           3       0.90      0.86      0.88      1000\n",
      "           4       0.77      0.77      0.77      1000\n",
      "           5       0.99      0.84      0.91      1000\n",
      "           6       0.64      0.59      0.61      1000\n",
      "           7       0.88      0.96      0.92      1000\n",
      "           8       0.98      0.93      0.95      1000\n",
      "           9       0.90      0.96      0.93      1000\n",
      "\n",
      "    accuracy                           0.85     10000\n",
      "   macro avg       0.86      0.85      0.85     10000\n",
      "weighted avg       0.86      0.85      0.85     10000\n",
      "\n"
     ]
    }
   ],
   "source": [
    "pred=model.best_estimator_.predict(x_test_scale)\n",
    "r=classification_report(y_test,pred)  #class report\n",
    "print('class report：', r, sep='\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 311
    },
    "id": "UNitS4rsI9tN",
    "outputId": "17b198ec-2ca0-4d9b-fa10-9efe515e1787"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "cnf_matrix = confusion_matrix(y_test,pred) \n",
    "np.set_printoptions(precision=2)\n",
    "\n",
    "class_names = range(10)\n",
    "plt.figure()\n",
    "plot_confusion_matrix(cnf_matrix   \n",
    "                      , classes=class_names\n",
    "                      , title='Confusion matrix')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Decision Tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Cr3rIIi5KFRZ",
    "outputId": "4df9ceb3-1594-4ddd-cc43-3c6309e336c8"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'max_depth': 10}\n"
     ]
    }
   ],
   "source": [
    "parms_grid={\n",
    "    'max_depth':[1,5,10,20]\n",
    "}\n",
    "dc =DecisionTreeClassifier()\n",
    "model = GridSearchCV(dc,parms_grid, cv=5)\n",
    "model.fit(x_train_scale, y_train)\n",
    "\n",
    "print(model.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "OS_C_2-XPQGf",
    "outputId": "d3940e2f-0d47-4dc9-b043-e4f5609ef34c"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'mean_fit_time': array([ 3.14,  9.26, 18.12, 30.19]),\n",
       " 'std_fit_time': array([1.2 , 0.22, 0.46, 0.81]),\n",
       " 'mean_score_time': array([0.04, 0.03, 0.04, 0.04]),\n",
       " 'std_score_time': array([0., 0., 0., 0.]),\n",
       " 'param_max_depth': masked_array(data=[1, 5, 10, 20],\n",
       "              mask=[False, False, False, False],\n",
       "        fill_value='?',\n",
       "             dtype=object),\n",
       " 'params': [{'max_depth': 1},\n",
       "  {'max_depth': 5},\n",
       "  {'max_depth': 10},\n",
       "  {'max_depth': 20}],\n",
       " 'split0_test_score': array([0.2 , 0.71, 0.81, 0.8 ]),\n",
       " 'split1_test_score': array([0.2, 0.7, 0.8, 0.8]),\n",
       " 'split2_test_score': array([0.2 , 0.71, 0.81, 0.8 ]),\n",
       " 'split3_test_score': array([0.2 , 0.71, 0.81, 0.81]),\n",
       " 'split4_test_score': array([0.2 , 0.71, 0.81, 0.8 ]),\n",
       " 'mean_test_score': array([0.2 , 0.71, 0.81, 0.8 ]),\n",
       " 'std_test_score': array([0., 0., 0., 0.]),\n",
       " 'rank_test_score': array([4, 3, 1, 2], dtype=int32)}"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.cv_results_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "PMoObeYgKMLK",
    "outputId": "d55e6903-4ffa-4e69-9195-c7482b524441"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "class report：\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.79      0.77      0.78      1000\n",
      "           1       0.97      0.92      0.95      1000\n",
      "           2       0.65      0.67      0.66      1000\n",
      "           3       0.82      0.81      0.81      1000\n",
      "           4       0.60      0.73      0.66      1000\n",
      "           5       0.93      0.88      0.90      1000\n",
      "           6       0.57      0.50      0.53      1000\n",
      "           7       0.87      0.93      0.90      1000\n",
      "           8       0.93      0.90      0.91      1000\n",
      "           9       0.91      0.91      0.91      1000\n",
      "\n",
      "    accuracy                           0.80     10000\n",
      "   macro avg       0.80      0.80      0.80     10000\n",
      "weighted avg       0.80      0.80      0.80     10000\n",
      "\n"
     ]
    }
   ],
   "source": [
    "pred=model.best_estimator_.predict(x_test_scale)\n",
    "r=classification_report(y_test,pred)  #class report\n",
    "print('class report：', r, sep='\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 311
    },
    "id": "TCGxt9GpKOQw",
    "outputId": "a6eedb79-da8c-46d2-96af-383416db42be"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "cnf_matrix = confusion_matrix(y_test,pred)  \n",
    "np.set_printoptions(precision=2)\n",
    "\n",
    "class_names = range(10)\n",
    "plt.figure()\n",
    "plot_confusion_matrix(cnf_matrix   \n",
    "                      , classes=class_names\n",
    "                      , title='Confusion matrix')\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}