ki-dhbw/tasks/18 - decisions.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import sklearn\n",
    "from sklearn import tree\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.preprocessing import StandardScaler, PolynomialFeatures\n",
    "import tensorflow as tf\n",
    "from matplotlib import pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 784)\n",
      "(10000, 784)\n"
     ]
    }
   ],
   "source": [
    "(train_images, train_labels), (test_images, test_labels) = tf.keras.datasets.mnist.load_data()\n",
    "x_train = train_images.reshape((len(train_images), -1))\n",
    "print(x_train.shape)\n",
    "x_test = test_images.reshape((len(test_images), -1))\n",
    "print(x_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Text(0.5, 0.9166666666666666, 'x[350] <= 131.5\\ngini = 0.9\\nsamples = 60000\\nvalue = [5923, 6742, 5958, 6131, 5842, 5421, 5918, 6265, 5851\\n5949]'),\n",
       " Text(0.25, 0.75, 'x[568] <= 0.5\\ngini = 0.881\\nsamples = 39160\\nvalue = [5433.0, 623.0, 4772.0, 1312.0, 5353.0, 3266.0, 4775.0\\n5773.0, 3081.0, 4772.0]'),\n",
       " Text(0.375, 0.8333333333333333, 'True  '),\n",
       " Text(0.125, 0.5833333333333334, 'x[430] <= 0.5\\ngini = 0.839\\nsamples = 24084\\nvalue = [697.0, 414.0, 904.0, 847.0, 5133.0, 1668.0, 2771.0\\n5568.0, 1439.0, 4643.0]'),\n",
       " Text(0.0625, 0.4166666666666667, 'x[405] <= 2.5\\ngini = 0.705\\nsamples = 9667\\nvalue = [258.0, 404.0, 565.0, 424.0, 503.0, 503.0, 315.0, 4974.0\\n836.0, 885.0]'),\n",
       " Text(0.03125, 0.25, 'x[484] <= 0.5\\ngini = 0.475\\nsamples = 6717\\nvalue = [247, 71, 401, 93, 336, 92, 166, 4803, 18, 490]'),\n",
       " Text(0.015625, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.046875, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.09375, 0.25, 'x[516] <= 14.5\\ngini = 0.848\\nsamples = 2950\\nvalue = [11, 333, 164, 331, 167, 411, 149, 171, 818, 395]'),\n",
       " Text(0.078125, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.109375, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.1875, 0.4166666666666667, 'x[211] <= 28.5\\ngini = 0.788\\nsamples = 14417\\nvalue = [439, 10, 339, 423, 4630, 1165, 2456, 594, 603, 3758]'),\n",
       " Text(0.15625, 0.25, 'x[98] <= 0.5\\ngini = 0.582\\nsamples = 6308\\nvalue = [88.0, 1.0, 139.0, 64.0, 3810.0, 285.0, 1369.0, 224.0\\n103.0, 225.0]'),\n",
       " Text(0.140625, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.171875, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.21875, 0.25, 'x[156] <= 0.5\\ngini = 0.76\\nsamples = 8109\\nvalue = [351, 9, 200, 359, 820, 880, 1087, 370, 500, 3533]'),\n",
       " Text(0.203125, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.234375, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.375, 0.5833333333333334, 'x[435] <= 0.5\\ngini = 0.793\\nsamples = 15076\\nvalue = [4736.0, 209.0, 3868.0, 465.0, 220.0, 1598.0, 2004.0\\n205.0, 1642.0, 129.0]'),\n",
       " Text(0.3125, 0.4166666666666667, 'x[489] <= 22.5\\ngini = 0.403\\nsamples = 6018\\nvalue = [4596.0, 11.0, 476.0, 155.0, 36.0, 436.0, 198.0, 18.0\\n66.0, 26.0]'),\n",
       " Text(0.28125, 0.25, 'x[380] <= 2.5\\ngini = 0.255\\nsamples = 5184\\nvalue = [4462, 6, 108, 142, 10, 266, 141, 11, 17, 21]'),\n",
       " Text(0.265625, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.296875, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.34375, 0.25, 'x[320] <= 1.0\\ngini = 0.728\\nsamples = 834\\nvalue = [134.0, 5.0, 368.0, 13.0, 26.0, 170.0, 57.0, 7.0, 49.0\\n5.0]'),\n",
       " Text(0.328125, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.359375, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.4375, 0.4166666666666667, 'x[346] <= 0.5\\ngini = 0.77\\nsamples = 9058\\nvalue = [140.0, 198.0, 3392.0, 310.0, 184.0, 1162.0, 1806.0\\n187.0, 1576.0, 103.0]'),\n",
       " Text(0.40625, 0.25, 'x[348] <= 95.5\\ngini = 0.506\\nsamples = 4505\\nvalue = [12.0, 191.0, 3111.0, 238.0, 25.0, 122.0, 238.0, 145.0\\n392.0, 31.0]'),\n",
       " Text(0.390625, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.421875, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.46875, 0.25, 'x[655] <= 0.5\\ngini = 0.755\\nsamples = 4553\\nvalue = [128.0, 7.0, 281.0, 72.0, 159.0, 1040.0, 1568.0, 42.0\\n1184.0, 72.0]'),\n",
       " Text(0.453125, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.484375, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.75, 0.75, 'x[489] <= 26.5\\ngini = 0.821\\nsamples = 20840\\nvalue = [490, 6119, 1186, 4819, 489, 2155, 1143, 492, 2770\\n1177]'),\n",
       " Text(0.625, 0.8333333333333333, '  False'),\n",
       " Text(0.625, 0.5833333333333334, 'x[290] <= 34.5\\ngini = 0.757\\nsamples = 10274\\nvalue = [397.0, 327.0, 328.0, 4413.0, 251.0, 1892.0, 499.0, 243.0\\n1091.0, 833.0]'),\n",
       " Text(0.5625, 0.4166666666666667, 'x[486] <= 58.5\\ngini = 0.597\\nsamples = 6272\\nvalue = [162.0, 299.0, 302.0, 3866.0, 86.0, 618.0, 250.0, 54.0\\n475.0, 160.0]'),\n",
       " Text(0.53125, 0.25, 'x[296] <= 2.5\\ngini = 0.446\\nsamples = 5042\\nvalue = [70.0, 259.0, 53.0, 3696.0, 76.0, 555.0, 28.0, 53.0\\n101.0, 151.0]'),\n",
       " Text(0.515625, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.546875, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.59375, 0.25, 'x[656] <= 0.5\\ngini = 0.805\\nsamples = 1230\\nvalue = [92.0, 40.0, 249.0, 170.0, 10.0, 63.0, 222.0, 1.0\\n374.0, 9.0]'),\n",
       " Text(0.578125, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.609375, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.6875, 0.4166666666666667, 'x[297] <= 5.5\\ngini = 0.817\\nsamples = 4002\\nvalue = [235, 28, 26, 547, 165, 1274, 249, 189, 616, 673]'),\n",
       " Text(0.65625, 0.25, 'x[486] <= 58.0\\ngini = 0.598\\nsamples = 1653\\nvalue = [64.0, 25.0, 5.0, 150.0, 52.0, 1015.0, 152.0, 20.0\\n110.0, 60.0]'),\n",
       " Text(0.640625, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.671875, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.71875, 0.25, 'x[598] <= 0.5\\ngini = 0.83\\nsamples = 2349\\nvalue = [171.0, 3.0, 21.0, 397.0, 113.0, 259.0, 97.0, 169.0\\n506.0, 613.0]'),\n",
       " Text(0.703125, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.734375, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.875, 0.5833333333333334, 'x[234] <= 0.5\\ngini = 0.66\\nsamples = 10566\\nvalue = [93.0, 5792.0, 858.0, 406.0, 238.0, 263.0, 644.0, 249.0\\n1679.0, 344.0]'),\n",
       " Text(0.8125, 0.4166666666666667, 'x[402] <= 0.5\\ngini = 0.463\\nsamples = 7859\\nvalue = [57, 5690, 395, 151, 139, 227, 384, 60, 610, 146]'),\n",
       " Text(0.78125, 0.25, 'x[149] <= 2.5\\ngini = 0.285\\nsamples = 6612\\nvalue = [3.0, 5567.0, 268.0, 71.0, 15.0, 80.0, 93.0, 36.0\\n429.0, 50.0]'),\n",
       " Text(0.765625, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.796875, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.84375, 0.25, 'x[103] <= 1.5\\ngini = 0.868\\nsamples = 1247\\nvalue = [54.0, 123.0, 127.0, 80.0, 124.0, 147.0, 291.0, 24.0\\n181.0, 96.0]'),\n",
       " Text(0.828125, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.859375, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.9375, 0.4166666666666667, 'x[658] <= 0.5\\ngini = 0.783\\nsamples = 2707\\nvalue = [36.0, 102.0, 463.0, 255.0, 99.0, 36.0, 260.0, 189.0\\n1069.0, 198.0]'),\n",
       " Text(0.90625, 0.25, 'x[345] <= 18.0\\ngini = 0.783\\nsamples = 1048\\nvalue = [17, 45, 377, 43, 43, 7, 254, 56, 82, 124]'),\n",
       " Text(0.890625, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.921875, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.96875, 0.25, 'x[515] <= 84.5\\ngini = 0.616\\nsamples = 1659\\nvalue = [19.0, 57.0, 86.0, 212.0, 56.0, 29.0, 6.0, 133.0, 987.0\\n74.0]'),\n",
       " Text(0.953125, 0.08333333333333333, '\\n  (...)  \\n'),\n",
       " Text(0.984375, 0.08333333333333333, '\\n  (...)  \\n')]"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "clf = tree.DecisionTreeClassifier(max_depth=10)\n",
    "clf.fit(x_train, train_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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
      "text/plain": [
       "<Figure size 4000x2000 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(40,20))\n",
    "_ = tree.plot_tree(clf, filled=True, fontsize=9, max_depth=5)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "score on training 0.8995333333333333\n",
      "score on test 0.8659\n",
      "6\n",
      "6\n"
     ]
    }
   ],
   "source": [
    "print(f\"score on training {clf.score(x_train, train_labels)}\")\n",
    "print(f\"score on test {clf.score(x_test, test_labels)}\")\n",
    "prediction = clf.predict(x_test)\n",
    "\n",
    "print(prediction[22])\n",
    "print(test_labels[22])\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": ".venv",
   "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": "3.12.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}