13. Hyperparameter.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Hyperparameter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from sklearn import datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "digits = datasets.load_digits()\n",
    "X = digits.data\n",
    "y = digits.target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 666)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9888888888888889"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "\n",
    "knn_clf = KNeighborsClassifier(n_neighbors = 3)\n",
    "knn_clf.fit(X_train, y_train)\n",
    "knn_clf.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Find the best k"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "best_k = 4 \n",
      "best_score =  0.9916666666666667\n"
     ]
    }
   ],
   "source": [
    "best_score = 0.0\n",
    "besk_k = 1\n",
    "for k in range(1, 11):\n",
    "    knn_clf = KNeighborsClassifier(n_neighbors = k)\n",
    "    knn_clf.fit(X_train, y_train)\n",
    "    score = knn_clf.score(X_test, y_test)\n",
    "    if score > best_score:\n",
    "        best_score = score\n",
    "        best_k = k\n",
    "print('best_k =', best_k, '\\nbest_score = ', best_score)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "note:如果best_k落在边界上，则需要重新定义包括边界的k的取值范围"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 考虑距离？不考虑距离？"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "best_k = 4 \n",
      "best_score =  0.9916666666666667 \n",
      "best_method =  uniform\n"
     ]
    }
   ],
   "source": [
    "best_method = ''\n",
    "best_score = 0.0\n",
    "besk_k = 1\n",
    "for method in ['uniform', 'distance']:\n",
    "    for k in range(1, 11):\n",
    "        knn_clf = KNeighborsClassifier(n_neighbors = k,\\\n",
    "                                       weights = method )\n",
    "        knn_clf.fit(X_train, y_train)\n",
    "        score = knn_clf.score(X_test, y_test)\n",
    "        if score > best_score:\n",
    "            best_score = score\n",
    "            best_k = k\n",
    "            best_method = method\n",
    "            \n",
    "print('best_k =', best_k, '\\nbest_score = ', best_score\\\n",
    "     , '\\nbest_method = ', best_method)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Manhattan Distance：\n",
    "$\\sum_{i = 1}^N|X^{a}_i -X^{b}_i|$\n",
    "\n",
    "Euclidean distance：\n",
    "$\\sqrt{\\sum_{i = 1}^N{|X^{a}_i -X^{b}_i|^2}}$\n",
    "\n",
    "Minkowski Distance：\n",
    "$\\sqrt[n]{\\sum_{i = 1}^N{|X^{a}_i -X^{b}_i|^n}}$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 搜索Minkowski Distance相应的p"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "best_k = 5 \n",
      "best_score =  0.9888888888888889 \n",
      "best_p =  1\n",
      "CPU times: user 13.9 s, sys: 41.2 ms, total: 14 s\n",
      "Wall time: 14 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "best_p = -1\n",
    "best_score = 0.0\n",
    "besk_k = 1\n",
    "for i in range(1, 6):\n",
    "    for k in range(1, 11):\n",
    "        knn_clf = KNeighborsClassifier(n_neighbors = k,\\\n",
    "                            weights = 'distance',p = i )\n",
    "        knn_clf.fit(X_train, y_train)\n",
    "        score = knn_clf.score(X_test, y_test)\n",
    "        if score > best_score:\n",
    "            best_score = score\n",
    "            best_k = k\n",
    "            best_p = i\n",
    "            \n",
    "print('best_k =', best_k, '\\nbest_score = ', best_score\\\n",
    "     , '\\nbest_p = ', best_p)"
   ]
  }
 ],
 "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": "3.7.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}