notebooks/attack_decision_tree.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ART decision tree classifier attack"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This notebook shows how to compute adversarial examples on decision trees (as described in by Papernot et al. in https://arxiv.org/abs/1605.07277). Due to the structure of the decision tree, an adversarial example can be computed without any explicit gradients, only by traversing the learned tree structure.\n",
    "\n",
    "Consider the following simple decision tree for four dimensional data, where we go to the left if a condition is true:\n",
    "\n",
    "                    F1<3\n",
    "           \n",
    "              F2<5        F2>2\n",
    "                \n",
    "         F4>3     C1    F3<1     C3* \n",
    "             \n",
    "      C1     C2       C3    C1  \n",
    "      \n",
    "Given sample [4,4,1,1], the tree outputs C3 (as indicated by the star). To misclassify the sample, we walk one node up and explore the subtree on the left. We find the leaf outputting C1 and change the two features, obtaining [4,1.9,0.9,1]. In this implementation, we change only the features with wrong values, and specify the offset in advance."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Applying the attack"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.datasets import load_digits\n",
    "from matplotlib import pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "from art.attacks.evasion import DecisionTreeAttack\n",
    "from art.estimators.classification import SklearnClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0 1 2 3 4 5 6 7 8 9 0 1 2 3]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.colorbar.Colorbar at 0x7fa24585ca10>"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAASQAAAD4CAYAAABFcpsMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAASuklEQVR4nO3df6zddX3H8edrpVCLVMSqEMoGGiUhulm9wzEWpyCuIgGTmQkJTNlMtyU6mDgG08X9uWSLwz8MyQ0oLHQ6rZAZpyBRCSNBpK0VhBZF5qRQLBUVRKHtva/9cb7XXcvlnu899/v9ns/p9/VIvun9nnPu5/M+NH3z+fX9fGSbiIgS/Ma4A4iImJOEFBHFSEKKiGIkIUVEMZKQIqIYh7VR6OE6wqs4so2ix+rA2m6/07HHPtFZXY88fXRnda3atb+zurz/QGd1dekZnmafn9VyyvijtxzpHz8xU+uzW+959hbbG5ZTXx2tJKRVHMkbdWYbRY/V3j8+rdP6/vayz3RW1z9sPa+zul79wd2d1XXgsR91VleX7vJXl13Gj5+Y4Zu3/Gatz6447ntrl11hDa0kpIgon4FZZscdxq9JQoroKWP2u16XrStJSBE9lhZSRBTBmJnCHh1LQorosVmSkCKiAAZmkpAiohRpIUVEEQzszxhSRJTAuLguW61n2SRtkPSApAclXdF2UBHRAcNMzasrQxOSpBXAJ4C3A6cAF0g6pe3AIqJdg5Xa9a5hJH1S0h5J31ngvQ9JsqShj5/UaSGdCjxo+yHb+4DPAN09+BQRLREzNa8argOe8/CtpBOAs4Af1imkTkI6Hnh43v2u6rWDK94oaYukLft5tk7dETFGg0Ft1bqGlmXfDiy0PcW/ApdX1Q1VZ1B7oWieU7jtaWAaYI2OKWukLCKeY7AOqfYOJmslbZl3P139m39eks4FHrH9balePXUS0i7ghHn364BHa5UeEUWbrdH6qey1PVX3w5JWAx8G3raUeOp02e4GXiXpJEmHA+cDX1hKJRFRnrkWUkNjSAd7JXAS8G1JP2DQkNkm6djFfmloC8n2AUnvB24BVgCftH3fKBFGRDmMmGlpF2vb9wIvm7uvktKU7b2L/V6thZG2vwR8aTkBRkR5ltBlW5SkTwNvZjDWtAv4qO1rl1pOVmpH9JQR+7yimbLsC4a8f2KdcpKQInpqsDCyrIOHkpAiemzEAevWJCFF9JQtZpwWUkQUYjYtpIgowWBQu6wUUFY0EdGZDGpPuC5PkgU4/6ifdFbXVUf/vLO6/mvbLZ3V9YZ//KvO6gJYO31np/Ut10xD65CakoQU0VNtrtQeVRJSRI/NZpYtIkoweLg2CSkiCmDE/oYeHWlKElJET9lkYWRElEJZGBkRZTBpIUVEQTKoHRFFMGpsg7amJCFF9NTgGKSyUkCdk2uf90TKiJhkjR4U2Yg6HcjrWOBEyoiYbGawUrvO1ZU6p47cLunE9kOJiK6VtmNkY6kvR2lHTBZbjbWQFhrakfTPknZKukfSTZKOHlZOYwnJ9rTtKdtTKzmiqWIjoiWDQe0Vta4aruO5Qzu3Aq+x/dvAd4ErhxVS1iKEiOjQYE/tOtcwtm8Hnjjota/YPlDdfoPB6bWLKmvOLyI6MxjUrj2GtFbSlnn307anl1DdnwH/MexDQxNSUydSRkR5lrBSe6/tqVHqkPRh4ACwadhn68yyLXoiZURMpi5Wakt6D3AOcKZtD/t8umwRPdbmJv+SNgB/B/yh7V/U+Z0kpIiesmH/bDMJaaGhHQazakcAt0oC+Ibtv1ysnCSkiJ4adNmaSUjPM7Sz5LHmJKSIHittpXYSUkRPLXHavxNJSBG91VyXrSlJSBE9lj21G3bgjDd0Vtf5R23vrC6At284v7O6XnTPzs7q+pM7zuysrifWz3RWF8DaTmtbnsEsW45BiogCZAvbiChKumwRUYTMskVEUTLLFhFFsMWBJKSIKEW6bBFRhIwhRURRkpAioghZhxQRRZm4dUiSTgD+DTgWmGWwuffH2w4sItplw4GGNmhrSp0W0gHgMtvbJB0FbJV0q+37W44tIlo2cV0227uB3dXPT0naARwPJCFFTLCJH0OSdCKwHrhrgfc2AhsBVrG6gdAiom2e1IQk6YXA54FLbT958PvVoXHTAGt0zNDjTiJi/Eob1K41oiVpJYNktMn2je2GFBFdsAdjSHWuYSR9UtIeSd+Z99oxkm6V9L3qzxcPK2doQtLg/JJrgR22PzY0soiYEGJm9jdqXTVcB2w46LUrgK/afhXw1ep+UXVqOh24CDhD0vbqOrtOhBFRNlu1ruHl+HbgiYNePg+4vvr5euCdw8qpM8t2BxTW0YyIZVvis2xrJW2Zdz9djRsv5uXVLD22d0t62bBKslI7oq88GEeqaa/tqRajAWoOakfEoWkW1bpG9CNJxwFUf+4Z9gtJSBE95WYHtRfyBeA91c/vAf5z2C8kIUX0mF3vGkbSp4E7gZMl7ZL058A/AWdJ+h5wVnW/qIwhRfRYUyu1bV/wPG8t6RC+JKSInhq0fsqaQE9CiuixiX64NiIOLUuY9u/ExCekZ17S3Vf4yJ7XdlYXwOw9Ozutryt33/vKcYcQVNuPTOAGbRFxiCqsgZSEFNFbGdSOiKIU1kRKQorosbSQIqIIBmZnk5AiogQG0kKKiFJkHVJElCMJKSLKUG972i7VOUp7FXA7cET1+c22P9p2YBHRgQlsIT0LnGH759VxSHdI+rLtb7QcW0S0yeBJm2WzbeDn1e3K6iosr0bEaMpKSHUPilwhaTuDPXFvtb3gUdqStkjasp9nm44zItrgmldHaiUk2zO2XwesA06V9JoFPjNte8r21EqOaDrOiGjDJCakObZ/CtzGc0+ojIhJM7cwss7VkTpHab9U0tHVzy8A3gocmhv1RPRMU5v8N6XOLNtxwPWSVjBIYJ+1/cV2w4qITjQ0yybpb4D3MWh33QtcbPuZpZZTZ5btHmD9kiOMiOKpgdaPpOOBvwZOsf1LSZ8FzgeuW2pZWakd0VfNDlgfBrxA0n5gNfDoKIWUtaFuRHSo5oD2YFB77dyynuraOFeK7UeAfwF+COwGfmb7K6NElBZSRJ/VbyHttT210BuSXgycB5wE/BT4nKQLbd+w1HDSQoros9ma1+LeCvyP7cdt7wduBH5/lHDSQoroq+Y2aPsh8HuSVgO/ZHB89pZRCkpCiuixJmbZbN8laTOwDTgAfAuYHqWsJKSIPmtolq3akmjZ2xJlDCkiijHxLaRnXtxdTt1052md1QXwar7ZaX1dOexF+zqr68DPDu+srknURJetSROfkCJiRKaxR0eakoQU0WdpIUVEKdJli4hyJCFFRDGSkCKiBHK6bBFRksyyRUQp0kKKiHIkIUVEEQocQ6r93EV1WOS3JGWD/4hDRWHnsi2lhXQJsANY01IsEdExDd98rVN1j9JeB7wDuKbdcCKiz+p22a4CLmeRzSwlbZzbAHw/zzYSXES0rLAuW52Ta88B9tjeutjnbE/bnrI9tZIjGgswIlri/18cOezqSp0xpNOBcyWdDawC1ki6wfaF7YYWEa2btFk221faXmf7RAanUX4tySjiEFFYly3rkCJ6SkzoLNsc27fZPqetYCKiQw2OIUk6WtJmSTsl7ZA00n7PaSFF9Flz3bGPAzfbfpekw4HVoxSShBTRZw0kJElrgDcB7wWwvQ8Y6SSHHIMU0WNL6LKtnVtnWF0b5xXzCuBx4FPV42XXSDpylHiSkCL6rP4s2965dYbVNf9k2sOA1wNX214PPA1cMUo4SUgRfeXBLFuda4hdwC7bd1X3mxkkqCVLQoroswbWIdl+DHhY0snVS2cC948STga1I3qswcdCPgBsqmbYHgIuHqWQiU9Iq37S3cqu333t9zurC+BnHdZ12LEv76yud5+y6GORjfrsl/+gs7omUkMJyfZ2YGq55Ux8QoqIEXX8WEgdSUgRPSXK28I2CSmix5KQIqIcSUgRUYwkpIgoQoHHICUhRfRZElJElKK0DdqSkCJ6LF22iCjDpC6MlPQD4ClgBjhge9lLxCOiAJOYkCpvsb23tUgiolNZqR0RRdFsWRmp7n5IBr4iaetBW1f+So7SjpgwdfdCKvBcttNtPyrpZcCtknbavn3+B6otLacB1uiYstJuRCyotC5brRaS7UerP/cANwGnthlURHSksBbS0IQk6UhJR839DLwN+E7bgUVE+5o6KLIpdbpsLwdukjT3+X+3fXOrUUVENwrrsg1NSLYfAn6ng1giokvOoyMRUYgS1yHlGKSIPrPrXTVIWlGdXPvFUcNJCymixxpuIV0C7ADWjFpAWkgRfdXgwkhJ64B3ANcsJ6S0kCJ6bAmD2mslbZl3P10thp5zFXA5cNRy4klCiuixJSSkvc+3y4ekc4A9trdKevNy4klCiugrU3vAeojTgXMlnQ2sAtZIusH2hUstaOIT0poHujtw+qPrRp48GMmfbvxgZ3WtfOfjndXVpZOuvHPcIRStiUFt21cCVwJULaQPjZKM4BBISBGxDIWtQ0pCiuipNhZG2r4NuG3U309Ciugru7gN2pKQIvqsrHyUhBTRZ6U9y5aEFNFXBtJli4hilJWPkpAi+ixdtogoRmbZIqIMBR6lXWv7EUlHS9osaaekHZJOazuwiGjXYGGka11dqdtC+jhws+13STocWN1iTBHRlUnbU1vSGuBNwHsBbO8D9rUbVkR0ocvWTx11umyvAB4HPlXtl3tNdT7br8lR2hETpsCjtOskpMOA1wNX214PPA1ccfCHbE/bnrI9tZIjGg4zIpo3eJatztWVOglpF7DL9l3V/WYGCSoiJl2Dp440YWhCsv0Y8LCkk6uXzgTubzWqiGhfdVBknasrdWfZPgBsqmbYHgIubi+kiOhMYYPatRKS7e3Aght8R8QEKysfZaV2RJ9ptqyFSElIEX1lJm9hZEQcmkS3j4XUkaO0I/qsgWl/SSdI+nr1nOt9ki4ZNZy0kCL6rJkW0gHgMtvbJB0FbJV0q+0lLw9KQoroq4bGkGzvBnZXPz8laQdwPCOsV0xCiuixJcyyrZW0Zd79tO3p55QnnQisB+46+L06kpAiemtJj4Xstb3oWkRJLwQ+D1xq+8lRIpr4hDR7z87O6nr31Zd1VhfARy77dGd1XfX9Mzur6+7XreisrliEaWyltqSVDJLRJts3jlrOxCekiFiGBsaQJAm4Fthh+2PLKSvT/hE91tAWtqcDFwFnSNpeXWePEk9aSBF91kCXzfYdDLboXrYkpIi+smGmrGdHkpAi+qywR0eSkCL6LAkpIopgICfXRkQZDM4YUkSUwBQ3qD10HZKkk+etLdgu6UlJl3YRXES0rLBTR4a2kGw/ALwOQNIK4BHgppbjioguTPig9pnA923/bxvBRESXum391LHUhHQ+sOATn5I2AhsBVrF6mWFFROsMFLbJf+1n2aoz2c4FPrfQ+zlKO2ICTdoY0jxvB7bZ/lFbwURElyb70ZELeJ7uWkRMIIMncR2SpNXAWcBftBtORHRqEldq2/4F8JKWY4mIrk34LFtEHCrs4mbZkpAi+iwtpIgog/HMzLiD+DVJSBF9le1HIqIohU3759SRiJ4y4FnXuoaRtEHSA5IelHTFqDElIUX0lasN2upci6h2AfkEg6c5TgEukHTKKCGlyxbRYw0Nap8KPGj7IQBJnwHOA+5fakFyC9N+kh4HlrpFyVpgb+PBlOFQ/W75XuPzW7ZfupwCJN3M4LvWsQp4Zt79tO3pqpx3ARtsv6+6vwh4o+33LzWmVlpIo/yHkrTF9lQb8Yzbofrd8r0mm+0NDRW10CGRI7V0MoYUEcu1Czhh3v064NFRCkpCiojluht4laSTqn3Tzge+MEpBJQ1qT487gBYdqt8t3yuwfUDS+4FbgBXAJ23fN0pZrQxqR0SMIl22iChGElJEFKOIhNTUsvOSSDpB0tcl7ZB0n6RLxh1TkyStkPQtSV8cdyxNknS0pM2SdlZ/d6eNO6Y+GfsYUrXs/LsMtsjdxWDE/gLbS17lWRJJxwHH2d4m6ShgK/DOSf9ecyR9EJgC1tg+Z9zxNEXS9cB/276mmjFabfun446rL0poIf1q2bntfcDcsvOJZnu37W3Vz08BO4DjxxtVMyStA94BXDPuWJokaQ3wJuBaANv7koy6VUJCOh54eN79Lg6Rf7hzJJ0IrAfuGm8kjbkKuBwoa++K5XsF8Djwqao7eo2kI8cdVJ+UkJAaW3ZeIkkvBD4PXGr7yXHHs1ySzgH22N467lhacBjweuBq2+uBp4FDYkxzUpSQkBpbdl4aSSsZJKNNtm8cdzwNOR04V9IPGHSvz5B0w3hDaswuYJftuZbsZgYJKjpSQkJqbNl5SSSJwVjEDtsfG3c8TbF9pe11tk9k8Hf1NdsXjjmsRth+DHhY0snVS2cywhYaMbqxPzrS5LLzwpwOXATcK2l79drf2/7SGGOK4T4AbKr+5/gQcPGY4+mVsU/7R0TMKaHLFhEBJCFFREGSkCKiGElIEVGMJKSIKEYSUkQUIwkpIorxfxL4fOXne+jCAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "digits = load_digits()\n",
    "X = digits.data\n",
    "y = digits.target\n",
    "\n",
    "clf = DecisionTreeClassifier()\n",
    "clf.fit(X,y)\n",
    "clf_art = SklearnClassifier(clf)\n",
    "print(clf.predict(X[:14]))\n",
    "plt.imshow(X[0].reshape(8,8))\n",
    "plt.colorbar()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We now craft adversarial examples and plot their classification. The difference is really small, and often only one or two features are changed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Decision tree attack: 100%|██████████| 14/14 [00:00<00:00, 1546.08it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[6 4 4 6 6 4 1 2 4 4 6 4 6 4]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7fa24570fa10>"
      ]
     },
     "execution_count": 3,
     "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": [
    "attack = DecisionTreeAttack(clf_art)\n",
    "adv = attack.generate(X[:14])\n",
    "print(clf.predict(adv))\n",
    "plt.imshow(adv[0].reshape(8,8))\n",
    "# plt.imshow((X[0]-adv[0]).reshape(8,8))  ##use this to plot the difference"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The change is possibly larger if we specify which class the sample should be (mis-)classified as. To do this, we just specify a label for each attack point."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Decision tree attack: 100%|██████████| 14/14 [00:00<00:00, 1073.48it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[6 6 7 7 8 8 9 9 1 1 2 2 3 3]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7fa245684f50>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPUAAAD4CAYAAAA0L6C7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAALLElEQVR4nO3df6jV9R3H8derq+U0zS1bhVdmjZK1jWVdHOIIptvQJRVssCtUrA0uDIoiWdRobPs/wv0xArFakEuaFURrtVhFBc78kXPZtWHS8Gal0fplpN587497BGu33e855/vj3PeeD7h4fxzu533Qp99zv/ec78cRIQB5nNT0AADKRdRAMkQNJEPUQDJEDSQzpYpverJPiWmaUcW3btTonHrv01lnvV3bWq8dml3bWtNGjta2VhwdrW2tOn2kQzoShz3e1yqJeppm6JteVsW3btRbP1hc63o/X72htrV+ue3y2tY6/8bXa1tr9I03a1urTpvjr5/5NR5+A8kQNZAMUQPJEDWQDFEDyRA1kAxRA8kQNZAMUQPJFIra9nLbL9veY/vmqocC0LkJo7bdJ+l3klZIukDSKtsXVD0YgM4UOVIvkrQnIvZGxBFJGyTV90RhAG0pEvVcSftO+Hik9blPsD1ke6vtrUd1uKz5ALSpSNTjvbzrv65WGBFrI2IgIgam6pTuJwPQkSJRj0iad8LH/ZL2VzMOgG4ViXqLpPNsn2P7ZEmDkh6udiwAnZrwIgkRMWr7WkmPS+qTdFdE7Kp8MgAdKXTlk4h4VNKjFc8CoAQ8owxIhqiBZIgaSIaogWSIGkiGqIFkiBpIppIdOrKqc8cMSRqc+e/a1loz+4Pa1vrT9sdrW+viX/+strUkac7aTbWuNx6O1EAyRA0kQ9RAMkQNJEPUQDJEDSRD1EAyRA0kQ9RAMkQNJFNkh467bB+w/WIdAwHoTpEj9e8lLa94DgAlmTDqiHhG0ts1zAKgBKW9Ssv2kKQhSZqm6WV9WwBtKu1EGdvuAL2Bs99AMkQNJFPkV1r3SdokaYHtEds/rX4sAJ0qspfWqjoGAVAOHn4DyRA1kAxRA8kQNZAMUQPJEDWQDFEDyUz6bXdGl15c21qDM3fUtpYkrVg+WNtap+3cXdtaK2YsqW2tt2/7uLa1JGlOrauNjyM1kAxRA8kQNZAMUQPJEDWQDFEDyRA1kAxRA8kQNZAMUQPJFLlG2TzbT9ketr3L9vV1DAagM0We+z0qaXVEbLc9U9I2209ExEsVzwagA0W23Xk9Ira33n9f0rCkuVUPBqAzbb1Ky/Z8SQslbR7na2y7A/SAwifKbJ8q6QFJN0TEe5/+OtvuAL2hUNS2p2os6PUR8WC1IwHoRpGz35Z0p6ThiLi9+pEAdKPIkXqJpKskLbW9o/X2/YrnAtChItvuPCfJNcwCoAQ8owxIhqiBZIgaSIaogWSIGkiGqIFkiBpIhqiBZCb9XlofnV7fXbj1wNdrW0uSjtW4v1Wddt/21aZHSI0jNZAMUQPJEDWQDFEDyRA1kAxRA8kQNZAMUQPJEDWQTJELD06z/bztv7e23flNHYMB6EyR51gelrQ0Ij5oXSr4Odt/joi/VTwbgA4UufBgSPqg9eHU1ltUORSAzhW9mH+f7R2SDkh6IiLG3XbH9lbbW4/qcNlzAiioUNQR8XFEXCipX9Ii218b5zZsuwP0gLbOfkfEO5KelrS8kmkAdK3I2e8zbM9uvf85Sd+RlPOFvkACRc5+ny3pHtt9GvtP4P6IeKTasQB0qsjZ750a25MawCTAM8qAZIgaSIaogWSIGkiGqIFkiBpIhqiBZIgaSGbyb7vz+fr+X1q/aXFta0nS+Xq+1vXqctpwff/s3v3KaG1r9QqO1EAyRA0kQ9RAMkQNJEPUQDJEDSRD1EAyRA0kQ9RAMkQNJFM46tYF/V+wzUUHgR7WzpH6eknDVQ0CoBxFt93pl3SppHXVjgOgW0WP1Gsk3STp2GfdgL20gN5QZIeOlZIORMS2/3U79tICekORI/USSZfZflXSBklLbd9b6VQAOjZh1BFxS0T0R8R8SYOSnoyIKyufDEBH+D01kExb15WJiKc1tpUtgB7FkRpIhqiBZIgaSIaogWSIGkiGqIFkiBpIZtJvuzNn56Ha1vry1a/VtpYkvVvjWlPOOrO2tVb+5Nna1rr/z9+qba1ewZEaSIaogWSIGkiGqIFkiBpIhqiBZIgaSIaogWSIGkiGqIFkCj1NtHUl0fclfSxpNCIGqhwKQOfaee73tyPircomAVAKHn4DyRSNOiT9xfY220Pj3YBtd4DeUPTh95KI2G/7i5KesL07Ip458QYRsVbSWkma5S9EyXMCKKjQkToi9rf+PCDpIUmLqhwKQOeKbJA3w/bM4+9L+p6kF6seDEBnijz8PlPSQ7aP3/4PEfFYpVMB6NiEUUfEXknfqGEWACXgV1pAMkQNJEPUQDJEDSRD1EAyRA0kQ9RAMpN+252TPjxS21q/6n+ktrUk6eqhG2tba+oVB2tbq07n3LKp6RFqx5EaSIaogWSIGkiGqIFkiBpIhqiBZIgaSIaogWSIGkiGqIFkCkVte7btjbZ32x62vbjqwQB0puhzv38r6bGI+KHtkyVNr3AmAF2YMGrbsyRdIunHkhQRRyTV9yoKAG0p8vD7XEkHJd1t+wXb61rX//4Ett0BekORqKdIukjSHRGxUNIhSTd/+kYRsTYiBiJiYKpOKXlMAEUViXpE0khEbG59vFFjkQPoQRNGHRFvSNpne0HrU8skvVTpVAA6VvTs93WS1rfOfO+VdE11IwHoRqGoI2KHpIGKZwFQAp5RBiRD1EAyRA0kQ9RAMkQNJEPUQDJEDSRD1EAyk34vrWM7d9e21o/uWF3bWpJ06+r7altrzSvLaltry4V9ta31/4gjNZAMUQPJEDWQDFEDyRA1kAxRA8kQNZAMUQPJEDWQzIRR215ge8cJb+/ZvqGO4QC0b8KniUbEy5IulCTbfZJek/RQxXMB6FC7D7+XSXolIv5VxTAAutfuCzoGJY37KgPbQ5KGJGka++cBjSl8pG5d8/sySX8c7+tsuwP0hnYefq+QtD0i3qxqGADdayfqVfqMh94AekehqG1Pl/RdSQ9WOw6AbhXddudDSadXPAuAEvCMMiAZogaSIWogGaIGkiFqIBmiBpIhaiAZogaScUSU/03tg5LafXnmHElvlT5Mb8h637hfzflSRJwx3hcqiboTtrdGxEDTc1Qh633jfvUmHn4DyRA1kEwvRb226QEqlPW+cb96UM/8TA2gHL10pAZQAqIGkumJqG0vt/2y7T22b256njLYnmf7KdvDtnfZvr7pmcpku8/2C7YfaXqWMtmebXuj7d2tv7vFTc/UrsZ/pm5tEPBPjV0uaUTSFkmrIuKlRgfrku2zJZ0dEdttz5S0TdIVk/1+HWf7RkkDkmZFxMqm5ymL7XskPRsR61pX0J0eEe80PVc7euFIvUjSnojYGxFHJG2QdHnDM3UtIl6PiO2t99+XNCxpbrNTlcN2v6RLJa1repYy2Z4l6RJJd0pSRByZbEFLvRH1XEn7Tvh4REn+8R9ne76khZI2NztJadZIuknSsaYHKdm5kg5Kurv1o8U62zOaHqpdvRC1x/lcmt+z2T5V0gOSboiI95qep1u2V0o6EBHbmp6lAlMkXSTpjohYKOmQpEl3jqcXoh6RNO+Ej/sl7W9ollLZnqqxoNdHRJbLKy+RdJntVzX2o9JS2/c2O1JpRiSNRMTxR1QbNRb5pNILUW+RdJ7tc1onJgYlPdzwTF2zbY39bDYcEbc3PU9ZIuKWiOiPiPka+7t6MiKubHisUkTEG5L22V7Q+tQySZPuxGa7G+SVLiJGbV8r6XFJfZLuiohdDY9VhiWSrpL0D9s7Wp/7RUQ82uBMmNh1kta3DjB7JV3T8Dxta/xXWgDK1QsPvwGUiKiBZIgaSIaogWSIGkiGqIFkiBpI5j/kuKMFZCbV2gAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "adv = attack.generate(X[:14],np.array([6,6,7,7,8,8,9,9,1,1,2,2,3,3]))\n",
    "print(clf.predict(adv))\n",
    "plt.imshow(adv[0].reshape(8,8))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, the attack has an offset parameter which specifies how close the new value of the feature is compared to the learned threshold of the tree. The default value is very small (0.001), however the value can be set larger when desired. Setting it to a very large value might however yield adversarial examples outside the range or normal features!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Decision tree attack: 100%|██████████| 14/14 [00:00<00:00, 1586.65it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[6 4 4 4 6 4 1 2 4 4 6 4 4 4]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.colorbar.Colorbar at 0x7fa24562fd90>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "attack = DecisionTreeAttack(clf_art,offset=20.0)\n",
    "adv = attack.generate(X[:14])\n",
    "print(clf.predict(adv))\n",
    "plt.imshow(adv[0].reshape(8,8))\n",
    "plt.colorbar()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "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": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}