recipes/ml_recipes_2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# K-Nearest Neighbours"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import datasets\n",
    "from sklearn import tree\n",
    "from sklearn.metrics import accuracy_score\n",
    "iris = datasets.load_iris()\n",
    "\n",
    "X = iris.data\n",
    "y = iris.target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.cross_validation import train_test_split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.973333333333\n"
     ]
    }
   ],
   "source": [
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "\n",
    "clf = KNeighborsClassifier()\n",
    "clf = clf.fit(X_train, y_train)\n",
    "\n",
    "predictions = clf.predict(X_test)\n",
    "print(accuracy_score(y_test, predictions))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Implement KNN from scratch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# first randomly, so it does no better than chance\n",
    "import random\n",
    "\n",
    "class ScrappyKNN:\n",
    "    def fit(self, X_train, y_train):\n",
    "        self.X_train = X_train\n",
    "        self.y_train = y_train\n",
    "        \n",
    "    def predict(self, X_test):\n",
    "        predictions = []\n",
    "        for row in X_test:\n",
    "            label = random.choice(self.y_train)\n",
    "            predictions.append(label)\n",
    "        return predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.4\n"
     ]
    }
   ],
   "source": [
    "clf = ScrappyKNN()\n",
    "clf.fit(X_train, y_train)\n",
    "predictions = clf.predict(X_test)\n",
    "print(accuracy_score(y_test, predictions))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Identifying Nearest Neighbour\n",
    "You can find the closest points using Euclidean distance. It works regardless of how many dimensions - you can just add more terms to the equation.\n",
    "\n",
    "## Euclidean Distance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.7320508075688772\n"
     ]
    }
   ],
   "source": [
    "x1, y1, z1 = (1,2,3)\n",
    "x2, y2, z2 = (2,3,4)\n",
    "e1 = (x1, y1, z1)\n",
    "e2 = (x2, y2, z2)\n",
    "\n",
    "import math\n",
    "def euclidean_distance(entity1, entity2):\n",
    "    summed_differences = 0\n",
    "    for i, feature in enumerate(entity1):\n",
    "        difference = math.pow((entity2[i] - entity1[i]), 2)\n",
    "        summed_differences += difference\n",
    "    return math.sqrt(summed_differences)\n",
    "\n",
    "d = euclidean_distance(e1, e2)\n",
    "print(d)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## SciPy Euclidean Implementation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.7320508075688772"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from scipy.spatial import distance\n",
    "distance.euclidean(e1, e2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=1 implementation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "class ScrappyKNN:\n",
    "    def fit(self, X_train, y_train):\n",
    "        self.X_train = X_train\n",
    "        self.y_train = y_train\n",
    "        \n",
    "    def predict(self, X_test):\n",
    "        predictions = []\n",
    "        for row in X_test:\n",
    "            label = self.closest(row)\n",
    "            predictions.append(label)\n",
    "        return predictions\n",
    "    \n",
    "    def closest(self, row):\n",
    "        best_dist = euclidean_distance(row, self.X_train[0])\n",
    "        best_index = 0\n",
    "        \n",
    "        for i in range(1, len(self.X_train)):\n",
    "            dist = euclidean_distance(row, self.X_train[i])\n",
    "            if dist < best_dist:\n",
    "                best_dist = dist\n",
    "                best_index = i\n",
    "                \n",
    "        return self.y_train[best_index]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.96\n"
     ]
    }
   ],
   "source": [
    "clf = ScrappyKNN()\n",
    "clf.fit(X_train, y_train)\n",
    "predictions = clf.predict(X_test)\n",
    "print(accuracy_score(y_test, predictions))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## KNN Pros/Cons\n",
    "\n",
    "- Pro:\n",
    "    - Simple\n",
    "- Con:\n",
    "    - Computationally intensive\n",
    "    - Hard to represent relationships between features\n",
    "    \n",
    "## k=n implementation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "class ScrappyKNN:\n",
    "    def __init__(self, k=1):\n",
    "        self.k = k\n",
    "    \n",
    "    def fit(self, X_train, y_train):\n",
    "        self.X_train = X_train\n",
    "        self.y_train = y_train\n",
    "        \n",
    "    def predict(self, X_test):\n",
    "        predictions = []\n",
    "        for row in X_test:\n",
    "            label = self.closest(row)\n",
    "            predictions.append(label)\n",
    "        return predictions\n",
    "    \n",
    "    def closest(self, row):\n",
    "        distances_idx = []\n",
    "        for i in range(len(self.X_train)):\n",
    "            dist = euclidean_distance(row, self.X_train[i])\n",
    "            distances_idx.append((dist, i))\n",
    "        distances_idx.sort()\n",
    "        \n",
    "        _, i = distances_idx[0]\n",
    "        labels_near = {self.y_train[i]: 1}\n",
    "        best_class = self.y_train[i]\n",
    "        num_best_class = 1\n",
    "        for i in range(1, self.k):\n",
    "            dist, index = distances_idx[i]\n",
    "            class_id = self.y_train[index]\n",
    "            if class_id not in labels_near:\n",
    "                labels_near[class_id] = 1\n",
    "            else:\n",
    "                labels_near[class_id] += 1\n",
    "            \n",
    "            if class_id == best_class:\n",
    "                num_best_class += 1\n",
    "            else:\n",
    "                if num_best_class < labels_near[class_id]:\n",
    "                    best_class = class_id\n",
    "                    num_best_class = labels_near[class_id]\n",
    "        return best_class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.973333333333\n"
     ]
    }
   ],
   "source": [
    "clf = ScrappyKNN(k=5)\n",
    "clf.fit(X_train, y_train)\n",
    "predictions = clf.predict(X_test)\n",
    "print(accuracy_score(y_test, predictions))"
   ]
  },
  {
   "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}