2.3.kNNtemplate-1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# k-Nearest Neighbour Classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load relevant packages \n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "np.random.seed(42)\n",
    "from sklearn import neighbors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load the data\n",
    "fullData = pd.read_csv('data/monthlyData.csv')\n",
    "\n",
    "# Make a small dataset with \"labels\"\n",
    "smallData = fullData[[\"ISO_3DIGIT\",\"Annual_temp\", \"Annual_precip\", \"tropical\",\"labels\"]]\n",
    "# What happens to the PLA when the training data is not linearly separable? \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "trainData = smallData[:150]\n",
    "testData = smallData[150:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x223655fd358>"
      ]
     },
     "execution_count": 4,
     "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": [
    "sns.scatterplot(x=\"Annual_temp\",y=\"Annual_precip\",hue=\"tropical\",data=smallData, palette=sns.color_palette(\"hls\", 2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Prepare training and test data\n",
    "xTrain = trainData.iloc[:,1:-2].values\n",
    "nTrain = xTrain.shape[0]\n",
    "yTrain = trainData.iloc[:,-1].values\n",
    "\n",
    "## Process test data\n",
    "xTest = testData.iloc[:,1:-2].values\n",
    "nTest = xTest.shape[0]\n",
    "\n",
    "yTest = testData.iloc[:,-1].values\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([False,  True, False, False, False, False, False, False, False,\n",
       "        True, False,  True,  True,  True, False,  True, False, False,\n",
       "        True,  True,  True,  True, False, False,  True, False, False,\n",
       "       False, False,  True,  True,  True,  True,  True,  True,  True,\n",
       "        True, False, False, False, False, False,  True, False,  True,\n",
       "       False, False, False, False, False,  True, False,  True, False,\n",
       "       False,  True, False, False,  True,  True,  True,  True,  True,\n",
       "        True, False,  True,  True,  True,  True, False,  True, False,\n",
       "        True,  True, False, False, False, False, False, False,  True,\n",
       "       False, False, False,  True, False,  True, False, False, False,\n",
       "        True, False,  True, False,  True, False, False, False, False,\n",
       "       False, False,  True,  True, False, False,  True, False, False,\n",
       "        True, False,  True,  True,  True, False,  True, False,  True,\n",
       "        True, False, False, False, False, False, False,  True,  True,\n",
       "        True,  True, False,  True, False, False,  True, False,  True,\n",
       "       False, False,  True, False, False, False, False,  True,  True,\n",
       "        True, False, False,  True,  True, False])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "yTrain"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### As k-NN is based on computing distances between neighbours, it is imperative to perform feature scaling!\n",
    "$X' = \\frac{(X-mean(X))}{std(X)} $"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 8.78535281 81.3418132 ] [ 17.81346667 115.03024667]\n",
      "[1. 1.] [-2.96059473e-18 -7.10542736e-17]\n"
     ]
    }
   ],
   "source": [
    "# Perform standard normalization on the training set. \n",
    "trMean = xTrain.mean(0)\n",
    "trStd = xTrain.std(0)\n",
    "print(trStd,trMean)\n",
    "xTrainRes = (xTrain-trMean)/trStd\n",
    "print(xTrainRes.std(0),xTrainRes.mean(0))\n",
    "# Use the obtained mean and standard deviation to standardize the test set\n",
    "xTestRes = (xTest-trMean)/trStd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x223658b9da0>"
      ]
     },
     "execution_count": 7,
     "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.scatter(xTrainRes[:,0],xTrainRes[:,1],c=yTrain)\n",
    "plt.scatter(xTestRes[0,0],xTestRes[0,1],marker='+')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Implement k-Nearest Neighbour classifier by completing this function\n",
    "\n",
    "# Functions such as np.argsort() can be useful here\n",
    "\n",
    "# Load the pairwise distance module\n",
    "from sklearn.metrics import pairwise_distances\n",
    "\n",
    "def kNN(X,Y,xTest,neighbors=1):\n",
    "    nTest = len(xTest)\n",
    "    yPred = np.zeros(nTest)\n",
    "    \n",
    "    # Compute pairwise distance between each test and training data point\n",
    "    dist = pairwise_distances(xTest,X)\n",
    "    \n",
    "    # Obtain the indices of the k-nearest neighbours\n",
    "    nNgbrs = np.argsort(dist)[:,:neighbors]\n",
    "    \n",
    "    # Obtain the votes of these k-NN\n",
    "    votes = Y[nNgbrs]\n",
    "    \n",
    "    # Predict the test set labels based on majority voting\n",
    "    for i in range(nTest):\n",
    "        yPred[i] = np.argmax(np.bincount(votes[i]))\n",
    "        \n",
    "    return yPred\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = xTrainRes[:4,:]\n",
    "Y = yTrain[:4]\n",
    "xT = xTestRes[:2,:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1.4058427  1.59946015 0.52827855 2.66554154]\n",
      " [1.34908728 2.31708954 1.22730134 2.95683334]]\n",
      "[[2 0 1 3]\n",
      " [2 0 1 3]]\n",
      "[[0 1 2 1]\n",
      " [0 1 2 1]]\n"
     ]
    }
   ],
   "source": [
    "dist = pairwise_distances(xT,X)\n",
    "print(dist)\n",
    "distSorted = np.argsort(dist)\n",
    "print(distSorted)\n",
    "votes = Y[distSorted]\n",
    "print(votes)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Accuracy of the KNN classifier is 0.892857\n"
     ]
    }
   ],
   "source": [
    "# Obtain predictions and print accuracy\n",
    "neighbors=3\n",
    "yPred = kNN(xTrain,yTrain,xTest,neighbors)\n",
    "acc = 1-(np.sum(yTest != yPred))/len(yTest)\n",
    "print('Test Accuracy of the KNN classifier is %f'%acc)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x223659b1ba8>"
      ]
     },
     "execution_count": 12,
     "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": [
    "# Plot the decision boundaries\n",
    "h = 0.1\n",
    "x_min, x_max = xTrain[:, 0].min() - 1, xTrain[:, 0].max() + 1\n",
    "y_min, y_max = xTrain[:, 1].min() - 1, xTrain[:, 1].max() + 1\n",
    "xx, yy = np.meshgrid(np.arange(x_min, x_max, h),\n",
    "                         np.arange(y_min, y_max, h))\n",
    "zInput = np.c_[xx.ravel(), yy.ravel()]\n",
    "Z = kNN(xTrain, yTrain, zInput,neighbors)\n",
    "Z = Z.reshape(xx.shape)\n",
    "plt.pcolormesh(xx, yy, Z,cmap=\"Spectral\")\n",
    "plt.scatter(xTrain[:,0],xTrain[:,1],c=yTrain,s=10)\n",
    "#plt.scatter(xTest[:,0],xTest[:,1],c=yPred,s=10)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Accuracy of the KNN classifier with 1 neighbours, is 0.857143\n",
      "Test Accuracy of the KNN classifier with 5 neighbours, is 0.928571\n",
      "Test Accuracy of the KNN classifier with 9 neighbours, is 0.857143\n",
      "Test Accuracy of the KNN classifier with 13 neighbours, is 0.857143\n",
      "Test Accuracy of the KNN classifier with 17 neighbours, is 0.857143\n",
      "Test Accuracy of the KNN classifier with 21 neighbours, is 0.821429\n",
      "Test Accuracy of the KNN classifier with 25 neighbours, is 0.785714\n",
      "Test Accuracy of the KNN classifier with 29 neighbours, is 0.785714\n",
      "Test Accuracy of the KNN classifier with 33 neighbours, is 0.785714\n",
      "Test Accuracy of the KNN classifier with 37 neighbours, is 0.785714\n",
      "Test Accuracy of the KNN classifier with 41 neighbours, is 0.785714\n",
      "Test Accuracy of the KNN classifier with 45 neighbours, is 0.785714\n",
      "Test Accuracy of the KNN classifier with 49 neighbours, is 0.785714\n"
     ]
    },
    {
     "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": [
    "numNgbrs = np.arange(1,52,4)\n",
    "testAcc = np.zeros(len(numNgbrs))\n",
    "\n",
    "for n in range(len(numNgbrs)):\n",
    "    yPred = kNN(xTrain,yTrain,xTest,numNgbrs[n])\n",
    "    testAcc[n] = 1-(np.sum(yTest != yPred))/len(yTest)\n",
    "    print('Test Accuracy of the KNN classifier with %d neighbours, is %f'%(numNgbrs[n],testAcc[n]))\n",
    "\n",
    "plt.plot(numNgbrs,testAcc)\n",
    "plt.grid()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 1,  3,  5,  7,  9, 11, 13, 15, 17, 19, 21])"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "numNgbrs = np.arange(1,22,2)\n",
    "testAcc = np.zeros(len(numNgbrs))\n",
    "numNgbrs"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Cross validation on the dataset\n",
    "### Task: Perform 5 fold cross validation and report aggregate test performance of the KNN classifier with k=[1,3,11,21]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 8.66949059 79.73725703] [ 17.92942857 113.42474286]\n",
      "[1. 1.] [-2.36636108e-16 -2.18238126e-16]\n"
     ]
    }
   ],
   "source": [
    "# Use 175 data points. Leave the last three out, for instance.\n",
    "\n",
    "smallData = smallData[:175]\n",
    "# Prepare training and test data\n",
    "xTrain = smallData.iloc[:,1:-2].values\n",
    "nTrain = xTrain.shape[0]\n",
    "\n",
    "yTrain = smallData.iloc[:,-1].values\n",
    "\n",
    "# Perform standard normalization on the training set. \n",
    "trMean = xTrain.mean(0)\n",
    "trStd = xTrain.std(0)\n",
    "print(trStd,trMean)\n",
    "xTrainRes = (xTrain-trMean)/trStd\n",
    "print(xTrainRes.std(0),xTrainRes.mean(0))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "35.0\n",
      "[0 1 2 3 4]\n",
      "[0 2 3 4]\n"
     ]
    }
   ],
   "source": [
    "kRange = np.asarray([1,3,5,11,21,31,51])\n",
    "ndata = len(smallData)\n",
    "nFold = 5\n",
    "nElFold = ndata/nFold\n",
    "print(nElFold)\n",
    "foldIdx = np.arange(ndata).reshape(-1,nFold)\n",
    "xT = xTrainRes[foldIdx[:,0],:]\n",
    "\n",
    "folds = np.arange(nFold)\n",
    "print(folds)\n",
    "print(np.delete(folds,1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Accuracy of the KNN classifier for fold 0, is 0.942857 with k=0\n",
      "Test Accuracy of the KNN classifier for fold 1, is 0.942857 with k=0\n",
      "Test Accuracy of the KNN classifier for fold 2, is 0.942857 with k=0\n",
      "Test Accuracy of the KNN classifier for fold 3, is 0.971429 with k=0\n",
      "Test Accuracy of the KNN classifier for fold 4, is 0.971429 with k=0\n",
      "Test Accuracy of the KNN classifier for fold 0, is 0.942857 with k=1\n",
      "Test Accuracy of the KNN classifier for fold 1, is 0.885714 with k=1\n",
      "Test Accuracy of the KNN classifier for fold 2, is 0.942857 with k=1\n",
      "Test Accuracy of the KNN classifier for fold 3, is 0.942857 with k=1\n",
      "Test Accuracy of the KNN classifier for fold 4, is 0.942857 with k=1\n",
      "Test Accuracy of the KNN classifier for fold 0, is 0.942857 with k=2\n",
      "Test Accuracy of the KNN classifier for fold 1, is 0.885714 with k=2\n",
      "Test Accuracy of the KNN classifier for fold 2, is 0.942857 with k=2\n",
      "Test Accuracy of the KNN classifier for fold 3, is 0.914286 with k=2\n",
      "Test Accuracy of the KNN classifier for fold 4, is 0.942857 with k=2\n",
      "Test Accuracy of the KNN classifier for fold 0, is 0.942857 with k=3\n",
      "Test Accuracy of the KNN classifier for fold 1, is 0.914286 with k=3\n",
      "Test Accuracy of the KNN classifier for fold 2, is 0.914286 with k=3\n",
      "Test Accuracy of the KNN classifier for fold 3, is 0.942857 with k=3\n",
      "Test Accuracy of the KNN classifier for fold 4, is 0.885714 with k=3\n",
      "Test Accuracy of the KNN classifier for fold 0, is 0.914286 with k=4\n",
      "Test Accuracy of the KNN classifier for fold 1, is 0.914286 with k=4\n",
      "Test Accuracy of the KNN classifier for fold 2, is 0.885714 with k=4\n",
      "Test Accuracy of the KNN classifier for fold 3, is 0.942857 with k=4\n",
      "Test Accuracy of the KNN classifier for fold 4, is 0.857143 with k=4\n",
      "Test Accuracy of the KNN classifier for fold 0, is 0.885714 with k=5\n",
      "Test Accuracy of the KNN classifier for fold 1, is 0.885714 with k=5\n",
      "Test Accuracy of the KNN classifier for fold 2, is 0.857143 with k=5\n",
      "Test Accuracy of the KNN classifier for fold 3, is 0.885714 with k=5\n",
      "Test Accuracy of the KNN classifier for fold 4, is 0.828571 with k=5\n",
      "Test Accuracy of the KNN classifier for fold 0, is 0.885714 with k=6\n",
      "Test Accuracy of the KNN classifier for fold 1, is 0.857143 with k=6\n",
      "Test Accuracy of the KNN classifier for fold 2, is 0.857143 with k=6\n",
      "Test Accuracy of the KNN classifier for fold 3, is 0.828571 with k=6\n",
      "Test Accuracy of the KNN classifier for fold 4, is 0.828571 with k=6\n"
     ]
    }
   ],
   "source": [
    "testAcc = np.zeros((len(kRange),len(folds)))\n",
    "for k in range(len(kRange)):\n",
    "    for fold in folds:\n",
    "        xT = xTrainRes[foldIdx[:,fold],:]\n",
    "        yT = yTrain[foldIdx[:,fold]].reshape(-1)\n",
    "        X = xTrainRes[foldIdx[:,np.delete(folds,fold)],:].reshape(-1,2)\n",
    "        Y = yTrain[foldIdx[:,np.delete(folds,fold)]].reshape(-1)\n",
    "        yPred = kNN(X,Y,xT,kRange[k])\n",
    "        testAcc[k,fold] = 1-(np.sum(yT != yPred))/len(yT)\n",
    "        print('Test Accuracy of the KNN classifier for fold %d, is %f with k=%d'%(fold,testAcc[k,fold],k));\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.94285714 0.94285714 0.94285714 0.97142857 0.97142857]\n",
      " [0.94285714 0.88571429 0.94285714 0.94285714 0.94285714]\n",
      " [0.94285714 0.88571429 0.94285714 0.91428571 0.94285714]\n",
      " [0.94285714 0.91428571 0.91428571 0.94285714 0.88571429]\n",
      " [0.91428571 0.91428571 0.88571429 0.94285714 0.85714286]]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([0.95428571, 0.93142857, 0.92571429, 0.92      , 0.90285714])"
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(testAcc)\n",
    "testAcc.mean(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7ff795e44860>]"
      ]
     },
     "execution_count": 121,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAD8CAYAAACb4nSYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAIABJREFUeJzt3Xl8VPW5x/HPkz2EkJBMCEgCCSQoyCaEgBpBBRWt16Va61pRK3axtXbV23tv77X1drna1t5Sq1VEaau1drPV1gUXNgWCLMqaBARCgCSEhCVk/90/ZuSmEWWASc5k5vt+vXgxc+Y3M8/B8Ttnfuec55hzDhERiQ4xXhcgIiI9R6EvIhJFFPoiIlFEoS8iEkUU+iIiUUShLyISRRT6IiJRRKEvIhJFFPoiIlEkzusCuvL5fC4vL8/rMkREepWVK1fWOueyjjUu7EI/Ly+P0tJSr8sQEelVzGxbMOM0vSMiEkUU+iIiUUShLyISRRT6IiJRRKEvIhJFFPoiIlFEoS8iEkUiJvTrG1t46NUy3tvZ4HUpIiJhK+xOzjpRMTHGz14ro6W9ndGD07wuR0QkLEXMln6/pHjOyE1n4eZar0sREQlbERP6AFNHZPFeVQN7DzZ7XYqISFiKuNB3DhaXa2tfRORoIir0xwxOI71PvKZ4REQ+QkSFfmyMUVLgY1FZDc45r8sREQk7ERX64J/iqT7QzMbdB7wuRUQk7ERe6Bf6ryGwcHONx5WIiISfiAv9gWlJnJqdysIyhb6ISFcRF/oAU0f4WLF1H40tbV6XIiISViI09LNoae9g2ZY6r0sREQkrERn6k/IySIqP4U3N64uI/JOIDP2k+Fgm52dqXl9EpIuIDH3wT/FsqTlE5b5Gr0sREQkbERv600b4AHR2rohIJxEb+sOz+jIoLUnH64uIdBKxoW9mTC3MYklFLW3tHV6XIyISFiI29ME/r3+gqY3VO+q9LkVEJCxEdOiXFPiIMbVkEBH5QESHflqfeMblpvNmmXbmiohAhIc++Buwra2sZ9+hFq9LERHxXOSHvq6mJSJyRMSH/ricNPolxWleX0SEKAj9uNgYSgp9LCqr1dW0RCTqBRX6ZjbTzDaZWbmZ3XOUx4ea2QIzW2tmb5hZTpfH+5nZTjP7eagKPx5TC7PYvb+JsuqDXry9iEjYOGbom1ksMAe4GBgFXGdmo7oMewB4yjk3FrgP+H6Xx78LvHny5Z6YqSN0NS0REQhuS78YKHfObXHOtQDPAJd3GTMKWBC4/Xrnx81sIpANvHzy5Z6YU9KTKRjQV62WRSTqBRP6g4Edne5XBpZ1tga4KnD7SiDVzDLNLAZ4EPjGyRZ6sqYWZrF8ax1Nre1elyIi4plgQt+OsqzrHtGvA9PMbBUwDdgJtAFfAF50zu3gY5jZbDMrNbPSmpru2RqfOsJHc1sHy7bqaloiEr3ighhTCeR2up8DVHUe4JyrAj4JYGZ9gauccw1mdiZwjpl9AegLJJjZQefcPV2e/yjwKEBRUVG3HGIzOT+ThLgYFm6uYVpgjl9EJNoEE/orgEIzy8e/BX8tcH3nAWbmA+qccx3AvcBcAOfcDZ3GzAKKugZ+T0lOiGVyfoZ25opIVDvm9I5zrg24E3gJ2AA865xbZ2b3mdllgWHnApvMbDP+nbb3d1O9J2VqYRZl1Qepqj/sdSkiIp6wcDthqaioyJWWlnbLa2/afYCLfrqQH141hk9PGtIt7yEi4gUzW+mcKzrWuIg/I7ezEdl9GdgvSZdQFJGoFVWhb2acU+hjcXkt7R3h9QtHRKQnRFXog//s3IbDrayp1NW0RCT6RF3olxT4MF1NS0SiVNSFfv+UBMYOTlPoi0hUirrQB/8Uz+od9TQ0tnpdiohIj4ra0O9wsKRCR/GISHSJytAfn5tOaqKupiUi0ScqQz8+NoazCjJZuLlGV9MSkagSlaEP/imeqoYmKmoOeV2KiEiPid7QL/R32nx1wx6PKxER6TlRG/q5GX2YlNefB1/exF9W7/S6HBGRHhG1oQ/w+KxJTBzan7ueWc1ji7Z4XY6ISLeL6tDvlxTPvFuKuWTMQL73wga+/+IGOtSTR0QiWDAXUYloSfGx/O91E/D1XccjC7dQc6CZH149lvjYqP4+FJEIFfWhDxAbY/zXZaczIDWRB17eTO2hFh6+YQIpifrnEZHIos3ZADPjzvML+eFVY1hcVsP1v3qbvQebvS5LRCSkFPpdfHrSEB69qYiNuw9w9S/fYkddo9cliYiEjEL/KGaMyua3t0+m7lALn3x4KeuqGrwuSUQkJBT6H2Hi0Aye+9yZxMUY1z7yNkvVnE1EIoBC/2MUZqfyxy+cxaD0JGbNXcELa3d5XZKIyElR6B/DoLRkfn/HWYzLTePOp9/h+3/fwLqqBjVqE5FeycItvIqKilxpaanXZXxIU2s733huLX9bW4VzcEpaEuePHMD0kdmcOSyTpPhYr0sUkShmZiudc0XHHKfQPz41B5p5fWM1r27Yw6KyWg63ttMnIZaSAh8zRmZz3mkDyEpN9LpMEYkyCv0e0NTazltb9rJgwx4WbKhmV0MTZjAuJ50ZgV8Bpw1Mxcy8LlVEIpxCv4c551i/az8LNlSzYMMe1lT6D/McnJ7M9MAXwJRhGSTGaRpIREJPoe+x6v1NvLaxmlc3VLO4vIam1g5SEmI5pzCL6SMHcN5pA/D11TSQiISGQj+MNLW2s7Sillc3VPPahmp27/dPA52Rm870kdnMGJnNiOy+mgYSkROm0A9TzjnWVe3n1cB+gHd3+qeBcvonM2NkNtNHDmByfiYJcTqaVkSCF9LQN7OZwENALPCYc+4HXR4fCswFsoA64EbnXKWZjQceBvoB7cD9zrnffdx7RXrod7Vnf9OR/QCLy2tpbuugb2IcU0f4mH6a/2igjJQEr8sUkTAXstA3s1hgM3ABUAmsAK5zzq3vNOb3wN+cc0+a2fnALc65m8xsBOCcc2VmdgqwEhjpnKv/qPeLttDv7HBLO0vKa1mw0f8roPpAM2YwYUh/po8cwIyR2RQO0DSQiHxYKEP/TOA/nXMXBe7fC+Cc+36nMeuAiwJb9wY0OOf6HeW11gBXO+fKPur9ojn0O+vo6DQNtHEP7+3cD8CQjD6cf5r/C6A4P0PTQCICBB/6wVwlZDCwo9P9SmBylzFrgKvwTwFdCaSaWaZzbm+ngoqBBKAiiPeMejExxpicNMbkpHH3BSPY3dB05BfA08u3M2/p+6QmxjH11CzumDqMsTnpXpcsIr1AMKF/tLmErj8Pvg783MxmAQuBnUDbkRcwGwTMB252znV86A3MZgOzAYYMGRJU4dFmYFoSN0weyg2Th9LY0saScv9JYf9Yt5sX1u7iE2MH8Y0LTyXPl+J1qSISxkIyvdNlfF9go3MuJ3C/H/AG8H3n3O+PVZCmd47PgaZWfrVwC79atJXW9g5umDyEL00v1DkAIlEm2OmdYCaEVwCFZpZvZgnAtcDzXd7MZ2YfvNa9+I/kITD+T8BTwQS+HL/UpHi+euGpvPmNc/n0pFx+vWw70370Og+9Wsah5rZjv4CIRJVjhr5zrg24E3gJ2AA865xbZ2b3mdllgWHnApvMbDOQDdwfWH4NMBWYZWarA3/Gh3olBAb0S+L+K8fw8t1TmToii5+8uplp//MG89/eRmv7h2bURCRK6eSsCPXO9n384MWNLH+/jnxfCt+46FQuHj1Qh3uKRKhQTu9ILzRhSH9+d8cUHr+5iPhY4wu/eYcrfrGUt7fsPfaTRSRiKfQjmJkxfWQ2f79rKj+6eizV+5u49tG3uXXeCjbu3u91eSLiAU3vRJGm1nbmLX2fOa+Xc7C5jasm5HD3BSMYnJ7sdWkicpLUcE0+Un1jC794o4J5S98H4Jaz8vjCuQWk9Yn3tjAROWEKfTmmyn2N/PiVzfxp1U5SE+P44nkF3HxWnq73K9ILKfQlaBt27eeH/9jIG5tqOCUtibsvGMEnJ+QQG6MjfUR6Cx29I0EbOagf824p5re3TyYrNZFvPLeWSx5axGsb9xBuGwUicnIU+nLEWcN9/PmLZzPn+gk0t7Vz67xSrn30bVZt3+d1aSISIgp9+SdmxifGDuKVr07ju5efTkXNQa78xVK+8JuVbKk56HV5InKSNKcvH+tgc1ugodsWmts6uK44ly9PL2RAapLXpYlIJ9qRKyFVc6CZny0o4+nl20mIi+Gz5wxj9tRh9E0Mpju3iHQ3hb50i621h3jgpU288O4uMlMS+PL0Qq4rHqIreIl4TEfvSLfI96Uw54YJ/PmLZ1OY3ZfvPL+OC37yJn9dU0VHR3htQIjIhyn05YSMz03n6dun8MQtk0iOj+VLT6/iil8sYWl5rdelicjHUOjLCTMzzjt1AC98+Rwe/NQ49h5s4frHlnHz3OWsr1JDN5FwpDl9CZmm1nbmv7WNn79ezv6mVq4cP5ivXjiCnP59vC5NJOJpR654pqGxlYffrOCJJVtxDj5z5lC+eF4B/VMSvC5NJGIp9MVzVfWH+emrm3luZSUpiXF8/tzh3Hp2vhq6iXQDhb6EjU27D/A/L23k1Q3VDOyXxN0XFHLVhBziYrVLSSRUdMimhI1TB6by2M2T+N3sKQxKT+Jbf3iXix9axCvr1dBNpKcp9KXHTB6WyR8/fxa/vHEC7R2O258q5ZpH3mLlNjV0E+kpCn3pUWbGzNGDeOnuqdx/5Wje39vIVQ8v5Y75pVSooZtIt9OcvniqsaWNxxdt5ZGFWzjc2s41RbncPaOQAf3U0E3keGhHrvQqew8287+vlfObZduIi4nhtpJ87pg2jNQkXbdXJBgKfemVtu09xIMvb+b5NVVkpCRw53kF3DBlCIlxOsxT5OPo6B3plYZmpvCz687gr3eWMHJQKvf9bT0zfvwmf1m9Uw3dREJAoS9haUxOGr++bTJP3VpMamI8dz2zmsvmLGZxmRq6iZwMhb6ELTNj6ogs/valEn766fHUN7Zy4+PL+Mzc5ew71OJ1eSK9kkJfwl5MjHHFGYNZ8LVp/NsnRvJ2xV5umbeCQ81tXpcm0uso9KXXSIyL5bPnDON/rz+DtZX1fO7XK2lua/e6LJFeJajQN7OZZrbJzMrN7J6jPD7UzBaY2Voze8PMcjo9drOZlQX+3BzK4iU6XXT6QH5w1VgWldXy1d+toV07eEWCdsyrWptZLDAHuACoBFaY2fPOufWdhj0APOWce9LMzge+D9xkZhnAd4AiwAErA8/VefdyUq4pyqW+sYX/fnEjaX3iuf+K0ZiZ12WJhL1gtvSLgXLn3BbnXAvwDHB5lzGjgAWB2693evwi4BXnXF0g6F8BZp582SIwe+pwPn/ucH67bDsPvLzJ63JEeoVgQn8wsKPT/crAss7WAFcFbl8JpJpZZpDPFTlh37zoVK4rzmXO6xU8tmiL1+WIhL1gQv9ov5m7TqJ+HZhmZquAacBOoC3I52Jms82s1MxKa2pqgihJxM/M+N4VY7h49EC+98IGnltZ6XVJImEtmNCvBHI73c8BqjoPcM5VOec+6Zw7A/h2YFlDMM8NjH3UOVfknCvKyso6zlWQaBcbY/z02vGUFPj41h/W8vK63V6XJBK2ggn9FUChmeWbWQJwLfB85wFm5jOzD17rXmBu4PZLwIVm1t/M+gMXBpaJhFRiXCyP3DSR0YPTuPPpVbxVsdfrkkTC0jFD3znXBtyJP6w3AM8659aZ2X1mdllg2LnAJjPbDGQD9weeWwd8F/8XxwrgvsAykZBLSYxj3qxJDMnow+1PlfLezgavSxIJO+qyKRFnV8Nhrn74LZpa2/n9585kWFZfr0sS6XbqsilRa1BaMvNvKwbgpseXs6vhsMcViYQPhb5EpGFZfXny1mIaDrdy0+PLqVODNhFAoS8RbPTgNB67uYjtdY3c8sRyDqpBm4hCXyLblGGZzLl+Au9V7eeO+aVq0CZRT6EvEe+CUdn88KqxLCnfy1eeWa0GbRLVFPoSFa6emMO/fWIkf39vN9/+07uE21FrIj3lmF02RSLFZ88Zxr7GFua8XkH/lAS+NfM0r0sS6XEKfYkqX7/wVPY1tvLwGxX07xPP7KnDvS5JpEcp9CWqmBnfvXw0DYdb+e8XN5KenMA1k3KP/USRCKHQl6gTG2P85Jrx7D/cyj1/XEu/5Hhmjh7odVkiPUI7ciUqJcTF8MsbJzIuN50vP72KpRW1Xpck0iMU+hK1UhLjeGLWJPJ8fbj9yVLWVtZ7XZJIt1PoS1RL75PAU7dOpn9KArOeWEF59UGvSxLpVgp9iXoD05KYf9tkYgw+8/gydtarQZtELoW+CJDvS2HeLcUcaGrjpseXsfdgs9cliXQLhb5IwAcN2nbuO8wt81aoQZtEJIW+SCeTAw3a1lXtZ/ZTpTS1qkGbRBaFvkgXM0Zl8z9Xj2VpxV7uemYVbe0dXpckEjIKfZGj+OSEHP7j0lG8tG4P/6oGbRJBdEauyEe4tSSf+sYWfvZaOf37JHDvJSO9LknkpCn0RT7G3ReMYF9jK48s3EL/lAQ+N00N2qR3U+iLfAwz478uO536w6384O8bSU+O59riIV6XJXLCFPoixxATYzz4qXHsP9zKv/7pXdKS47l4zCCvyxI5IdqRKxKEhLgYHr5xAuNz07nrmdUsLlODNumdFPoiQeqTEMfcWZPI96Uwe34pq3eoQZv0Pgp9keOQ3ieBp24rJrNvArc8sZzy6gNelyRyXBT6Iscpu18S82+dTGxMDDc+tpzKfY1elyQSNIW+yAnI86Xw1K3FHGpp4zOPL6dWDdqkl1Doi5ygUaf0Y+6sSVQ1HGbWE8s50NTqdUkixxRU6JvZTDPbZGblZnbPUR4fYmavm9kqM1trZpcElseb2ZNm9q6ZbTCze0O9AiJempSXwcM3TGTjrgPcrgZt0gscM/TNLBaYA1wMjAKuM7NRXYb9G/Csc+4M4FrgF4HlnwISnXNjgInAHWaWF5rSRcLDeacN4IFPjePtLXV86Wk1aJPwFsyWfjFQ7pzb4pxrAZ4BLu8yxgH9ArfTgKpOy1PMLA5IBlqA/SddtUiYueKMwfznv4zilfV7uOeP79LRoQZtEp6COSN3MLCj0/1KYHKXMf8JvGxmXwJSgBmB5c/h/4LYBfQB7nbO1Z1MwSLhatbZ+exrbOWhBWWkJ8fz7U+MxMy8LkvknwSzpX+0T23XzZjrgHnOuRzgEmC+mcXg/5XQDpwC5ANfM7NhH3oDs9lmVmpmpTU1Nce1AiLh5CszCrn5zKE8tngrD79Z4XU5Ih8STOhXArmd7ufw/9M3H7gNeBbAOfcWkAT4gOuBfzjnWp1z1cASoKjrGzjnHnXOFTnnirKyso5/LUTChJnxnX85ncvHn8KP/rGJ3y7b7nVJIv8kmNBfARSaWb6ZJeDfUft8lzHbgekAZjYSf+jXBJafb34pwBRgY6iKFwlHMTHGA58ax7mnZvHtP7/Li+/u8rokkSOOGfrOuTbgTuAlYAP+o3TWmdl9ZnZZYNjXgNvNbA3wNDDL+S81NAfoC7yH/8vjCefc2m5YD5GwEh8bw8M3TGTikP7c9cwqFpVp2lLCg4XbZeCKiopcaWmp12WIhETD4VY+/chbbK9r5DefncwZQ/p7XZJEKDNb6Zz70PR5VzojV6QbpSXH89Stxfj6JnLLvBVs3qMGbeIthb5INxvQL4lf3zaZ+NgYbnp8GTvq1KBNvKPQF+kBQzL7MP+2Yg63tPOZucupOaAGbeINhb5IDzltYD+euGUSuwIN2varQZt4QKEv0oMmDs3glzdOZNPuA3z2STVok56n0BfpYeeeOoAHrxnHivfruPO376hBm/Qohb6IBy4fP5j7LjudVzdU880/rFWDNukxwTRcE5FucNOZeexrbOXHr2wmPTmBf79UDdqk+yn0RTz0pfMLqDvUwtwlW8lIiefO8wu9LkkinEJfxENmxn9cOoqGw6088PJm0vskcOOUoV6XJRFMoS/isZgY40dXj2X/4Vb+/S/vkZYcz7+MO8XrsiRCaUeuSBiIj41hzg0TmDQ0g68+u5o3N6tBm3QPhb5ImEiKj+VXNxdRMCCVz81fycpt+7wuSSKQQl8kjHzQoC27XyK3zlvBpt1q0CahpdAXCTNZqYnMv20ySfFq0Cahp9AXCUO5GX146tbJNLd1cOPjy9SgTUJGoS8Spk4dmMrcWZOo3t/MNY+8xc9fK2PNjnradfaunARdOUskzC0tr+V7L2xg/a79gH/e/+yCTEoKsjin0EduRh+PK5RwEOyVsxT6Ir1E7cFmlpTXsrislsXltexqaAIgL7MPJYU+SgqyOHN4JmnJ8R5XKl5Q6ItEMOccFTUHWVTm/xJ4a8teGlvaiY0xxuWkUVLo/xUwPjed+FjN4kYDhb5IFGlp62DV9n0sLq9lUVktayvr6XDQNzGOKcMyOafQxzmFPvJ9KWrqFqEU+iJRrKGxlaUVtSwqr2VRWQ076g4DMDg9mZICHyWFPs4u8JGRkuBxpRIqCn0ROWLb3kNHpoKWVNRyoKkNMxh9SholhT7OKfAxMa8/iXGxXpcqJ0ihLyJH1dbewdqdDf4dwmW1vLN9H20djuT4WIrzMwJTQVmMyO6rqaBeRKEvIkE52NzG2xV7WVxey8KyGrbUHAJgQGrikamgkkIfA1KTPK5UPk6woa/WyiJRrm9iHDNGZTNjVDYAVfWHWVzm3x/w+qZq/rhqJwCnDUylpMDHOSOyKM7LIDlBU0G9kbb0ReQjdXQ41u/a798fUF7Diq37aGnvICE2hqK8/pQU+phamMWoQf2IidFUkJc0vSMiIXe4pZ3l79exuKyGRWW1bAx0Ac1ISeCs4f5DQ0sKsxicnuxxpdFH0zsiEnLJCbFMG5HFtBFZAFQfaGJJ4NyARWW1/G3tLgCGZaVwToH/C2DKsAxSk3SWcLjQlr6IhIRzjs17DrIo8Ctg2da9NLV2EBdjnDEknZKCLEoKfYzLSSNOZwmHXEind8xsJvAQEAs85pz7QZfHhwBPAumBMfc4514MPDYWeAToB3QAk5xzTR/1Xgp9kcjQ3NbOym37/DuFy2p5r6oB5yA1KY6zhmdSUpjF1EIfQzNTvC41IoQs9M0sFtgMXABUAiuA65xz6zuNeRRY5Zx72MxGAS865/LMLA54B7jJObfGzDKBeudc+0e9n0JfJDLVHWphaUXtkS+BnfX+s4RzM5KPdAw9a3gm6X10lvCJCOWcfjFQ7pzbEnjhZ4DLgfWdxjj8W/IAaUBV4PaFwFrn3BoA59ze4MoXkUiTkZLApWNP4dKxp+CcY2vtIf+5AZtr+euaKp5evp0YgzE56YH9AT4mDOlPQpymgkIpmC39q4GZzrnPBu7fBEx2zt3Zacwg4GWgP5ACzHDOrTSzrwATgQFAFvCMc+5HR3mP2cBsgCFDhkzctm1bKNZNRHqJ1vYO1uyoD+wQrmFNZQPtHY4+CbFMGZZJSYGPqSN8DM/SWcIfJZRb+kf7F+76TXEdMM8596CZnQnMN7PRgdcvASYBjcCCQGEL/unFnHsUeBT80ztB1CQiESQ+NoaivAyK8jK4+4IR7G9q5a2KvSwqq2FxWS2vbawGYGC/JH+voEDDOF/fRI8r732CCf1KILfT/Rz+f/rmA7cBMwGcc2+ZWRLgCzz3TedcLYCZvQhMABYgIvIR+iXFc9HpA7no9IEA7KhrDLSNruGV9Xt4bmUlAKMG9QucG+BjUl4GSfE6S/hYgpneicO/I3c6sBP/jtzrnXPrOo35O/A759w8MxuJP9QH4z+aZwH+rf0W4B/AT5xzL3zU+2lHroh8nPYOx3s7GwL7A2p4Z/s+WtsdiXExFOdn+FtFFGZx2sDUqDpLONSHbF4C/BT/4ZhznXP3m9l9QKlz7vnAETu/Avrin/r5pnPu5cBzbwTuDSx/0Tn3zY97L4W+iByPQ81tLN9ax8LAVFBZ9UEAfH0TOLvAd+RLYGBaZDeMUxsGEYlKuxuajkwFLSmvpfZgCwCFA/oe2R8wOT+TlMTIakig0BeRqNfR4di4+wCLy/1nCS/fWkdzWwfxscaEIf2P9AoaMziN2F4+FaTQFxHpoqm1ndL39x1pFbF+134A0pLjObsg88hJYrkZfTyu9Pgp9EVEjqH2YPORhnGLy2rZvd/fISYvs4//4jEFWZw5PJO05PBvGKfQFxE5Ds45KmoOsnBzLYvLa3l7y14aW9qJjTHG5aRRUuj/FTA+N534MGwYp9AXETkJLW0drNq+z3+WcHkt71bW0+H8VxqbMiwzcC1hH/m+lLA4S1ihLyISQg2NrSyt8H8BLCqrYUedv2Hc4PTkI9cSPrvAR0aKNw3jFPoiIt1o295DR3oFLa3Yy4GmNsxg9Clp/kNDC3xMzOtPYlzPnCWs0BcR6SFt7R2s3dkQaBtdw6rt9bR1OJLjYynOzwhMBWUxIrv7GsYp9EVEPHKwuY23Aw3jFpXXsqXmEAADUhOPTAWVFPgY0C90Zwkr9EVEwsTO+sMsKatlYeAs4X2NrQCcNjD1yJfA5PxMkhNOfCpIoS8iEoY6Ohzrd+0/sj+g9P19tLR3kBAbw4WnZ/Pz6yec0OuGsp++iIiESEyMMXpwGqMHp/H5c4dzuKWd5e/XsWhzDYnx3X/8v0JfRMRDyQmxTBuRxbQRWT3yfuF3WpmIiHQbhb6ISBRR6IuIRBGFvohIFFHoi4hEEYW+iEgUUeiLiEQRhb6ISBQJuzYMZlYDbAtiqA+o7eZywkm0rS9onaOF1jk0hjrnjnmGV9iFfrDMrDSYPhORItrWF7TO0ULr3LM0vSMiEkUU+iIiUaQ3h/6jXhfQw6JtfUHrHC20zj2o187pi4jI8evNW/oiInKcel3om9lMM9tkZuVmdo/X9XQHM5trZtVm9l6nZRlm9oqZlQX+7u9ljaFmZrlm9rqZbTCzdWZ2V2B5xK63mSWZ2XIzWxNY5/8KLM83s2WBdf6dmSV4XWsomVmsma0ys78F7kf6+r5vZu+a2WozKw0s8+xz3atC38xigTnAxcAo4DozG+VtVd1iHjCzy7J7gAXOuUJgQeB+JGkDvuacGwlMAb4Y+G8byevdDJzvnBsHjAeF8XWJAAACm0lEQVRmmtkU4IfATwLrvA+4zcMau8NdwIZO9yN9fQHOc86N73SYpmef614V+kAxUO6c2+KcawGeAS73uKaQc84tBOq6LL4ceDJw+0ngih4tqps553Y5594J3D6APxQGE8Hr7fwOBu7GB/444HzgucDyiFpnM8sBPgE8FrhvRPD6fgzPPte9LfQHAzs63a8MLIsG2c65XeAPSGCAx/V0GzPLA84AlhHh6x2Y6lgNVAOvABVAvXOuLTAk0j7jPwW+CXQE7mcS2esL/i/yl81spZnNDizz7HPd266Ra0dZpsOPIoiZ9QX+AHzFObffvyEYuZxz7cB4M0sH/gSMPNqwnq2qe5jZpUC1c26lmZ37weKjDI2I9e3kbOdclZkNAF4xs41eFtPbtvQrgdxO93OAKo9q6Wl7zGwQQODvao/rCTkzi8cf+L9xzv0xsDji1xvAOVcPvIF/f0a6mX2wQRZJn/GzgcvM7H38U7Pn49/yj9T1BcA5VxX4uxr/F3sxHn6ue1vorwAKA3v7E4Brgec9rqmnPA/cHLh9M/AXD2sJucDc7uPABufcjzs9FLHrbWZZgS18zCwZmIF/X8brwNWBYRGzzs65e51zOc65PPz/777mnLuBCF1fADNLMbPUD24DFwLv4eHnutednGVml+DfOogF5jrn7ve4pJAzs6eBc/F34tsDfAf4M/AsMATYDnzKOdd1Z2+vZWYlwCLgXf5/vvdf8c/rR+R6m9lY/DvxYvFvgD3rnLvPzIbh3xLOAFYBNzrnmr2rNPQC0ztfd85dGsnrG1i3PwXuxgG/dc7db2aZePS57nWhLyIiJ663Te+IiMhJUOiLiEQRhb6ISBRR6IuIRBGFvohIFFHoi4hEEYW+iEgUUeiLiESR/wO6RZ6kbYfYvwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(kRange,testAcc.mean(1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model selection using nested cross validation\n",
    "## Hyper-parameter of the k-NN algorithm is the \"k\". How do we choose k? \n",
    "* Split the data into say, 5 folds. \n",
    "* Use 3 folds for training and one fold for validation. \n",
    "* Try different k=[1,3,11,21]\n",
    "* Choose the model with best validation error\n",
    "* Report test set on the last fold.\n",
    "\n",
    "Repeat this process for each configuration of folds.\n"
   ]
  },
  {
   "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": "3.7.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}