2016-fall/seminars/sem04-linclass.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Машинное обучение, ФКН ВШЭ\n",
    "\n",
    "# Семинар 4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Populating the interactive namespace from numpy and matplotlib\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING: pylab import has clobbered these variables: ['plt', 'random']\n",
      "`%matplotlib` prevents importing * from pylab and numpy\n"
     ]
    }
   ],
   "source": [
    "%pylab inline\n",
    "import matplotlib.pylab as plt\n",
    "import math\n",
    "import random\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.cross_validation import train_test_split"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Линейная классификация\n",
    "\n",
    "Ранее мы изучили задачу восстановления регрессии и способы её решения. В частности, для линейной модели с функционалом качества MSE решение данной задачи выписывается аналитически через матрицу \"объект-признак\" и вектор ответов обучающей выборки.\n",
    "\n",
    "Давайте приступим к рассмотрению задачи классификации. Начнём с того, что вспомним её постановку.\n",
    "\n",
    "### Постановка задачи классификации\n",
    "\n",
    "Пусть задана обучающая выборка $X = \\left\\{ \\left( x_i, y_i \\right) \\right\\}_{i=1}^l, x_i \\in \\mathbb{X}, y_i \\in \\mathbb{Y},$ — $l$ пар объект-ответ, где\n",
    "$\\mathbb{X}$ — пространство объектов,\n",
    "$\\mathbb{Y}$ — пространство ответов.\n",
    "\n",
    "Задача классификации отличается о задачи восстановления регрессии тем, что в данном случае пространство ответов конечно:\n",
    "$$\\left| \\mathbb{Y} \\right| < \\infty.$$\n",
    "\n",
    "\n",
    "**Примеры**:\n",
    "\n",
    "1) Кредитный скоринг.\n",
    "* Объект — заявка на кредит от клиента банка.\n",
    "* Ответ — решение о выдаче или невыдаче кредита.\n",
    "\n",
    "2) Медицинская диагностика.\n",
    "* Объект — характеристики пациента (результаты анализов, общая информация о пациенте, история болезни).\n",
    "* Ответ — диагноз (одно из ограниченного множества заболеваний).\n",
    "\n",
    "В дальнейшем будем рассматривать задачу бинарной классификации (то есть разделения объектов на 2 группы) и полагать $\\mathbb{Y} = \\{-1, +1\\}.$\n",
    "\n",
    "### Линейная модель\n",
    "\n",
    "Сегодня мы будем рассматривать линейную модель. В случае классификации формула для ответа на объекте $x$ видоизменяется по сравнению с линейной регрессией (при этом мы предполагаем, что нулевой признак в выборке — константный):\n",
    "$$a(x) = sign(w_0 \\cdot 1 + w_1 x_1 + \\dots w_d x_d) = sign(\\sum_{j = 0}^d w_j x_j) = sign(\\langle w, x \\rangle).$$\n",
    "\n",
    "Множество точек $\\langle w, x\\rangle = 0$ является гиперплоскостью и делит пространство на 2 части по разные стороны от этой самой гиперплоскости. При этом объекты, расположенные по разные стороны от неё, полученный классификатор относит к различным классам в зависимости от знака скалярного произведения.\n",
    "\n",
    "<tr>\n",
    "<td> <img src=\"images/plane.png\" alt=\"Drawing\" style=\"width: 350px;\"/> </td>\n",
    "<td> <img src=\"images/cube.png\" alt=\"Drawing\" style=\"width: 350px;\"/> </td>\n",
    "</tr>\n",
    "\n",
    "Отметим, что для некоторого объекта $x$ расстояние до гиперплоскости для фиксированного вектора весов $w$ можно записать как $\\frac{\\left| \\langle w, x \\rangle \\right|}{\\|w \\|}$.\n",
    "\n",
    "Поэтому чем больше $\\left| \\langle w, x \\rangle \\right|$, тем дальше объект отстоит от гиперплоскости, тем увереннее алгоритм в его классификации.\n",
    "\n",
    "### Оптимизируемый функционал\n",
    "\n",
    "Как и в случае линейной регрессии, поиск оптимального алгоритма среди множества всех линейных моделей в случае классификации производится путем оптимизации некоторого функционала. Можно было бы рассматривать MAE или MSE, как и в случае регрессии, однако в случае классификации этот подход кажется немного странным, как минимум, потому что классы могут задаваться своими названиями, а не номерами. Их можно было бы занумеровать, но тогда результат будет сильно зависеть от способа нумерации.\n",
    "\n",
    "Будем минимизировать долю неверных классификаций, которую можно записать в следующем виде:\n",
    "\n",
    "$$Q(w, X) = \\frac{1}{l} \\sum_{i = 1}^l [a(x_i) \\ne y_i] \\to \\min_w.$$\n",
    "\n",
    "В этом функционале слагаемое $[a(x_i) \\ne y_i]$ отвечает за то, была ли допущена алгоритмом $a(\\cdot)$ ошибка на объекте $x_i$.\n",
    "\n",
    "Заметим, что $[a(x_i) \\ne y_i] = [sign \\, \\langle w, x_i \\rangle \\ne sign \\, y_i] = [y_i \\langle w, x_i \\rangle < 0],$ поэтому функционал можно переписать в следующем виде:\n",
    "$$Q(w, X) = \\frac{1}{l} \\sum_{i = 1}^l [y_i \\langle w, x_i \\rangle  < 0] \\to \\min_w.$$\n",
    "\n",
    "Величина $M_i = y_i \\langle w, x_i \\rangle$ называется **отступом** объекта $x_i$. Как было показано ранее:\n",
    "* объект верно классифицируется тогда и только тогда, когда отступ положительный;\n",
    "* чем больше значение отступа, тем надёжнее классифицируется объект.\n",
    "\n",
    "Тем не менее, получившийся функционал не получится оптимизировать градиентными методами. Поэтому рассмотрим пороговую функцию потерь для одного объекта $L(M) = [M < 0],$ участвующую в формуле функционала, оценим её сверху некоторой гладкой функцией потерь и будем минимизировать её:\n",
    "\n",
    "$$Q(w, X) = \\frac{1}{l} \\sum_{i = 1}^l [M_i < 0] = \\frac{1}{l} \\sum_{i = 1}^l L(M_i) \\le \\frac{1}{l} \\sum_{i = 1}^l \\widetilde{L}(M_i) = \\widetilde{Q}(w, X) \\to \\min_{w}$$\n",
    "\n",
    "Примеры оценок сверху для одного объекта:\n",
    "<img src=\"images/est.png\" alt=\"Drawing\" style=\"width: 650px;\">\n",
    "\n",
    "**ВАЖНО**: мы минимизируем верхнюю оценку и надеемся, что тогда будет минимизироваться и исходный функционал $Q(w, X)$, однако нет никаких гарантий, что так и будет!\n",
    "\n",
    "В зависимости от выбранной верхней оценки будут меняться свойства итогового алгоритма, положение разделяющей гиперплоскости и т.п.\n",
    "\n",
    "### Итог\n",
    "\n",
    "Таким образом, процесс обучения линейного классификатора складывается из следующих шагов:\n",
    "\n",
    "* Получение обучающей выборки $X = \\{ \\left( x_i, y_i \\right) \\}_{i=1}^l$\n",
    "* Выбор верхней оценки для пороговой функции потерь — отсюда получаем общий вид оптимизируемого функционала $\\widetilde{Q}(w, X^l)$\n",
    "* Проводим оптимизацию при помощи некоторого метода оптимизации и получаем оптимальное значение $w^*$\n",
    "* Предсказываем ответы для новых объектов по формуле $a(x) = \\langle w^*, x\\rangle$\n",
    "\n",
    "### Логистическая регрессия\n",
    "\n",
    "Рассмотрим в качестве верхней оценки пороговой функции потерь логистическую функцию:\n",
    "\n",
    "$$\\widetilde{L}(M) = \\log_2(1 + \\exp(-M)).$$\n",
    "\n",
    "Таким образом, необходимо решить следующую оптимизационную задачу:\n",
    "$$\\frac{1}{l} \\sum_{i=1}^l \\widetilde{L} (M_i) = \\frac{1}{l} \\sum_{i=1}^l \\log_2 (1 + \\exp (-y_i \\langle w, x_i \\rangle)) \\to \\min_w$$\n",
    "\n",
    "Получившийся метод обучения называется **логистической регрессией**.\n",
    "\n",
    "Одно из полезных свойств логистической регрессии, которое будет изучено нами несколько позднее, — тот факт, что она позволяет предсказывать помимо метки класса ещё и вероятность принадлежности каждому из них, что может быть полезным в некоторых задачах.\n",
    "\n",
    "**Пример**: Вы работаете в банке и хотите выдавать кредиты только тем клиентам, которые вернут его с вероятностью не меньше 0.9.\n",
    "\n",
    "### Пример обучения логистической регрессии\n",
    "#### Определение спама по тексту электронного письма\n",
    "\n",
    "Попробуем при помощи моделей линейной классификации построить алгоритм, отделяющий спам от нормальной почты. Для экспериментов воспользуемся небольшим набором данных с [UCI](https://archive.ics.uci.edu/ml/datasets.html). Объекты в датасете соответствуют письмам, которые описаны признаками на основе текста письма, спам — положительный пример для классификации, хорошее письмо — отрицательный пример.\n",
    "\n",
    "![Spam](./images/spam.jpg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>word_freq_make</th>\n",
       "      <th>word_freq_address</th>\n",
       "      <th>word_freq_all</th>\n",
       "      <th>word_freq_3d</th>\n",
       "      <th>word_freq_our</th>\n",
       "      <th>word_freq_over</th>\n",
       "      <th>word_freq_remove</th>\n",
       "      <th>word_freq_internet</th>\n",
       "      <th>word_freq_order</th>\n",
       "      <th>word_freq_mail</th>\n",
       "      <th>...</th>\n",
       "      <th>char_freq_;</th>\n",
       "      <th>char_freq_(</th>\n",
       "      <th>char_freq_[</th>\n",
       "      <th>char_freq_!</th>\n",
       "      <th>char_freq_$</th>\n",
       "      <th>char_freq_#</th>\n",
       "      <th>capital_run_length_average</th>\n",
       "      <th>capital_run_length_longest</th>\n",
       "      <th>capital_run_length_total</th>\n",
       "      <th>spam</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.64</td>\n",
       "      <td>0.64</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.32</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.778</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3.756</td>\n",
       "      <td>61</td>\n",
       "      <td>278</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.21</td>\n",
       "      <td>0.28</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.14</td>\n",
       "      <td>0.28</td>\n",
       "      <td>0.21</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.94</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.132</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.372</td>\n",
       "      <td>0.180</td>\n",
       "      <td>0.048</td>\n",
       "      <td>5.114</td>\n",
       "      <td>101</td>\n",
       "      <td>1028</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.06</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.71</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.23</td>\n",
       "      <td>0.19</td>\n",
       "      <td>0.19</td>\n",
       "      <td>0.12</td>\n",
       "      <td>0.64</td>\n",
       "      <td>0.25</td>\n",
       "      <td>...</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.143</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.276</td>\n",
       "      <td>0.184</td>\n",
       "      <td>0.010</td>\n",
       "      <td>9.821</td>\n",
       "      <td>485</td>\n",
       "      <td>2259</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.63</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.63</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.63</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.137</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.137</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3.537</td>\n",
       "      <td>40</td>\n",
       "      <td>191</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.63</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.63</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.63</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.135</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.135</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3.537</td>\n",
       "      <td>40</td>\n",
       "      <td>191</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 58 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   word_freq_make  word_freq_address  word_freq_all  word_freq_3d  \\\n",
       "0            0.00               0.64           0.64           0.0   \n",
       "1            0.21               0.28           0.50           0.0   \n",
       "2            0.06               0.00           0.71           0.0   \n",
       "3            0.00               0.00           0.00           0.0   \n",
       "4            0.00               0.00           0.00           0.0   \n",
       "\n",
       "   word_freq_our  word_freq_over  word_freq_remove  word_freq_internet  \\\n",
       "0           0.32            0.00              0.00                0.00   \n",
       "1           0.14            0.28              0.21                0.07   \n",
       "2           1.23            0.19              0.19                0.12   \n",
       "3           0.63            0.00              0.31                0.63   \n",
       "4           0.63            0.00              0.31                0.63   \n",
       "\n",
       "   word_freq_order  word_freq_mail  ...   char_freq_;  char_freq_(  \\\n",
       "0             0.00            0.00  ...          0.00        0.000   \n",
       "1             0.00            0.94  ...          0.00        0.132   \n",
       "2             0.64            0.25  ...          0.01        0.143   \n",
       "3             0.31            0.63  ...          0.00        0.137   \n",
       "4             0.31            0.63  ...          0.00        0.135   \n",
       "\n",
       "   char_freq_[  char_freq_!  char_freq_$  char_freq_#  \\\n",
       "0          0.0        0.778        0.000        0.000   \n",
       "1          0.0        0.372        0.180        0.048   \n",
       "2          0.0        0.276        0.184        0.010   \n",
       "3          0.0        0.137        0.000        0.000   \n",
       "4          0.0        0.135        0.000        0.000   \n",
       "\n",
       "   capital_run_length_average  capital_run_length_longest  \\\n",
       "0                       3.756                          61   \n",
       "1                       5.114                         101   \n",
       "2                       9.821                         485   \n",
       "3                       3.537                          40   \n",
       "4                       3.537                          40   \n",
       "\n",
       "   capital_run_length_total  spam  \n",
       "0                       278     1  \n",
       "1                      1028     1  \n",
       "2                      2259     1  \n",
       "3                       191     1  \n",
       "4                       191     1  \n",
       "\n",
       "[5 rows x 58 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas\n",
    "import urllib2\n",
    "\n",
    "SPAMBASE_NAMES_URL = 'https://archive.ics.uci.edu/ml/machine-learning-databases/spambase/spambase.names'\n",
    "SPAMBASE_DATA_URL = 'https://archive.ics.uci.edu/ml/machine-learning-databases/spambase/spambase.data'\n",
    "\n",
    "feature_names = [\n",
    "    line.strip().split(':')[0] \n",
    "    for line in urllib2.urlopen(SPAMBASE_NAMES_URL).readlines()[33:]\n",
    "]\n",
    "spam_data = pandas.read_csv(SPAMBASE_DATA_URL, header=None, names=(feature_names + ['spam']))\n",
    " \n",
    "X, y = spam_data.ix[:, :-1].values, spam_data.ix[:, -1].values\n",
    " \n",
    "spam_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Обучение логистической регрессии"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Разделим выборку на обучающую и тестовую в отношении 80/20 и обучим логистическую регрессию при помощи объекта [LogisticRegression](http://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "# splitting data\n",
    "train_part = 0.8\n",
    "n_train = int(train_part * X.shape[0])\n",
    "X_tr = X[:n_train]\n",
    "X_test = X[n_train:]\n",
    "y_tr = y[:n_train]\n",
    "y_test = y[n_train:]\n",
    "\n",
    "# training\n",
    "lr = LogisticRegression()\n",
    "lr = lr.fit(X_tr, y_tr)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Оценим долю верных прогнозов полученной модели (accuracy) при помощи соответствующей функции из модуля [sklearn.metrics](http://scikit-learn.org/stable/modules/classes.html#module-sklearn.metrics)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy = 0.813246471227\n"
     ]
    }
   ],
   "source": [
    "from sklearn import metrics\n",
    "\n",
    "preds = lr.predict(X_test)\n",
    "\n",
    "print 'Accuracy =', metrics.accuracy_score(y_test, preds)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Сгенерируем двумерную искуственную выборку из 2 различных нормальных распределений:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABIMAAAJPCAYAAADv8RQCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\npIUcywAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11220bf10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "mean0 = [10, 5]\n",
    "cov0 = [[1, 0], [0, 5]]  # diagonal covariance\n",
    "data0 = np.random.multivariate_normal(mean0, cov0, 1000)\n",
    "mean1 = [0, 0]\n",
    "cov1 = [[3, 1], [0, 1]]\n",
    "data1 = np.random.multivariate_normal(mean1, cov1, 1000)\n",
    "data = np.vstack((data0, data1))\n",
    "y_vis = np.hstack((-np.ones(1000), np.ones(1000)))\n",
    "\n",
    "\n",
    "plt.scatter(data0[:, 0], data0[:, 1], c='red')\n",
    "plt.scatter(data1[:, 0], data1[:, 1], c='green')\n",
    "plt.legend(['y = -1', 'y = 1'])\n",
    "axes = plt.gca()\n",
    "axes.set_xlim([-5,15])\n",
    "axes.set_ylim([-5,10])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Обучим логистическую регрессию:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
       "          intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1,\n",
       "          penalty='l2', random_state=None, solver='liblinear', tol=0.0001,\n",
       "          verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train_vis, X_test_vis, y_train_vis, y_test_vis = train_test_split(data, y_vis, test_size=0.2)\n",
    "\n",
    "from sklearn import linear_model\n",
    "\n",
    "logreg = linear_model.LogisticRegression(penalty = 'l2')\n",
    "logreg.fit(X_train_vis, y_train_vis)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Полученные в результате оптимизации коэффициенты линейной модели содержатся в атрибутах coef\\_ и intercept\\_ соответствующего объекта. Визуализируем разделяющую гиперплоскость алгоритма и рассмотрим значения предсказанных моделью вероятностей принадлежности нового объекта каждому из классов в зависимости от его координат."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(array([[-1.09187855, -0.36095293]]), array([ 6.2919995]))\n"
     ]
    }
   ],
   "source": [
    "print(logreg.coef_, logreg.intercept_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABIMAAAJPCAYAAADv8RQCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x110ea6c90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "w_1 = logreg.coef_[0][0]\n",
    "w_2 = logreg.coef_[0][1]\n",
    "w_0 = logreg.intercept_[0]\n",
    "\n",
    "\n",
    "plt.figure(figsize=(20,10))\n",
    "plt.scatter(data0[:, 0], data0[:, 1], c='red')\n",
    "plt.scatter(data1[:, 0], data1[:, 1], c='green')\n",
    "plt.legend(['y = -1', 'y = 1'])\n",
    "x_arr = np.linspace(-10, 15, 3000)\n",
    "plt.plot(x_arr, -(w_0 + w_1 * x_arr) / w_2)\n",
    "axes = plt.gca()\n",
    "axes.set_xlim([-5,15])\n",
    "axes.set_ylim([-5,10])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "P(y = -1|x) = 0.995268130627\n",
      "P(y = 1|x) = 0.00473186937303\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABIMAAAJPCAYAAADv8RQCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nGQQAAAAwIMogAAAAgAFRBgEAAAAMyL8BsO6krlQqtAEAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1121fba50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "point = np.array([[10, 2]]) # изменяем только координаты объекта\n",
    "\n",
    "plt.figure(figsize=(20,10))\n",
    "plt.scatter(data0[:, 0], data0[:, 1], c='red')\n",
    "plt.scatter(data1[:, 0], data1[:, 1], c='green')\n",
    "plt.scatter(point[:, 0], point[:, 1], marker = '*', s = 300, color = 'magenta')\n",
    "plt.legend(['y = -1', 'y = 1'])\n",
    "x_arr = np.linspace(-10, 15, 3000)\n",
    "plt.plot(x_arr, -(w_0 + w_1 * x_arr) / w_2)\n",
    "axes = plt.gca()\n",
    "axes.set_xlim([-5,15])\n",
    "axes.set_ylim([-5,10])\n",
    "prob = logreg.predict_proba(point)\n",
    "print 'P(y = -1|x) =', prob[0][0]\n",
    "print 'P(y = 1|x) =', prob[0][1]\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### SGDClassifier\n",
    "\n",
    "Объект [SGDClissifier](http://scikit-learn.org/stable/modules/generated/sklearn.linear_model.SGDClassifier.html) позволяет обучать линейные модели классификации и регрессии с помощью стохастического градиентного спуска.\n",
    "\n",
    "Полезные параметры:\n",
    "* loss - функция потерь (по факту то, какую модель обучаем): **hinge** (SVM), **log** (логистическая регрессия), **perceptron** (персептрон) и другие;\n",
    "* penalty - тип регуляризации: **l1**, **l2**, **elasticnet** (смесь l1 и l2 регуляризации);\n",
    "* alpha - коэффициент регуляризации;\n",
    "* fit_intercept - необходимо ли добавлять в модель свободный член (True/False);\n",
    "* n_iter - число эпох (полных проходов по выборке) при обучении;\n",
    "* learning_rate - шаг градиентного спуска (оптимизируется по умолчанию)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy = 0.86319218241\n"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import SGDClassifier\n",
    "lr_sgd = SGDClassifier(loss=\"log\", alpha=0.05, n_iter=200, fit_intercept=True)\n",
    "lr_sgd.fit(X_tr, y_tr)\n",
    "preds_sgd = lr_sgd.predict(X_test)\n",
    "print 'Accuracy =', metrics.accuracy_score(y_test, preds_sgd)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Оценка качества работы классификатора\n",
    "### Бинарные метрики\n",
    "Обучение и оценка качества модели производится на независимых множествах примеров. Как правило, имеющующиеся примеры разбивают на два подмножества: обучение (`train`) и контроль (`test`). Выбор пропорции разбиения — компромисс: большой размер обучения ведет к более богатым информацией и качественным алгоритмам, большой размер контрольной выборки ведет к менее шумной оценке качества.\n",
    "\n",
    "Для оценки качества классификации рассматривают [матрицу ошибок](http://en.wikipedia.org/wiki/Confusion_matrix): строчки соответствуют прогнозу модели, столбцы — истинным ответам, ячейки содержат число примеров тестовой выборки. Если для некоторого алгоритма $a(\\cdot)$ и объекта $x$ выполняется $a(x) = 1$, то говорят, что алгоритм $a$ _выделяет_ объект $x$.\n",
    "\n",
    "| | y = +1 | y = -1 |\n",
    "|------|------|\n",
    "|   **a(x) = +1**  | TP| FP |\n",
    "| **a(x) = -1**| FN | TN |\n",
    "\n",
    "Диагональ матрицы ошибок содержит правильно классифицированные положительные (TP) и отрицательные (TN) примеры. False Positive (FP) — ошибки I рода (ложное срабатывание, положили в спам хорошее письмо), False Negative (FN) — ошибки II рода (не отфильтровали спам). Ошибки I и II рода могут иметь различную стоимость.\n",
    "\n",
    "Часто рассматриваются следующие метрики качества бинарной классификации:\n",
    "  - Доля правильных ответов (Accuracy):\n",
    "  $$accuracy = \\frac{TP + TN}{TP + TN + FP + FN}$$\n",
    "  - Точность/precision (доля действительно положительных объектов среди объектов, выделенных алгоритмом):\n",
    "  $$precision = \\frac{TP}{TP + FP}$$\n",
    "  - Полнота/recall (доля выделенных алгоритмом объектов среди всех положительных объектов выборки):\n",
    "  $$recall = \\frac{TP}{TP + FN}$$\n",
    "  - $F_1$-мера (среднее гармоническое между точностью и полнотой)\n",
    "  $$F_1 = \\frac{2 \\cdot Precision \\cdot Recall}{Precision + Recall}$$\n",
    "  \n",
    "Подробнее про метрики качества бинарной классификации на Википедии: [Precision and Recall](http://en.wikipedia.org/wiki/Precision_and_recall).\n",
    "\n",
    "### Выбор порога классификации\n",
    "Многие модели классификации получают оценку принадлежности положительному классу $b(x) \\in \\mathbb{R}$, после чего принимается решение о классификации объекта путем сравнения оценки с некоторым порогом:\n",
    "\n",
    "$a(x) = \\begin{cases}\n",
    "        +1, \\, b(x) \\ge t,\\\\\n",
    "        -1, \\, b(x) < t.\n",
    "        \\end{cases}$\n",
    "        \n",
    "Матрица ошибок и все производные от нее метрики (Accuracy, Precision, Recall, etc.) зависят от порога $t$:\n",
    "![Порог классификации](./images/binary_threshold.png) \n",
    "\n",
    "Поскольку иногда необходимо оценить качество модели $b(x),$ предсказывающей не конкретную метку класса, а степень принадлежности классу (т.е. вне зависимости от значения порога), рассматривают координатную плоскость, по осям которой отложены значения метрик, а качество работы классификатора представлено в виде кривой — траектории изменения соответствующих метрик при варьировании порога:\n",
    "![Принцип построения ROC-кривой](./images/roc_trajectory.png)\n",
    "\n",
    "Наиболее распространены ROC-кривые и Precision/Recall кривые.\n",
    " - По осям Ox и Oy ROC-кривой отложены соответственно False Positive Rate (FPR) и True Positive Rate (TPR):\n",
    " $$FPR = \\frac{FP}{FP + TN},$$\n",
    " $$TPR = \\frac{TP}{FN + TP}.$$\n",
    " - По осям Ox и Oy PR-кривой отложены соответственно Recall и Precision.\n",
    "\n",
    "\n",
    "Аббревиатура ROC ([Receiver Operating Characteristic](en.wikipedia.org/wiki/Receiver_operating_characteristic)) была унаследована из инженерного дела.\n",
    "\n",
    "![Примеры ROC-кривых](./images/roc_and_pr_curves.png)\n",
    "\n",
    "В случае, если необходимо сравнить качество классификаторов вне зависимости от порога, применяют интегральные числовые метрики, например AUC-ROC (**A**rea **U**nder RO**C**) — площадь под ROC-кривой классификатора. AUC-ROC идеально работающего классификатора равно 1. Идеальный случайный классификатор в среднем имеет AUC-ROC=0.5.\n",
    "\n",
    "Построим описанные кривые для логистической регрессии, обученной на описанном выше датасете."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* PR-кривая проходит через точку (0,0).\n",
    "\n",
    "* Если при каком-то значении порога $t$ алгоритм $a(x)$ идеально разделяет объекты 2 классов, то PR-кривая проходит через точку (1,1).\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "from sklearn.cross_validation import train_test_split\n",
    "\n",
    "X_train_curve, X_test_curve, y_train_curve, y_test_curve = train_test_split(X, y, test_size=0.2)\n",
    "\n",
    "lr = LogisticRegression()\n",
    "lr = lr.fit(X_train_curve, y_train_curve)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from sklearn.metrics import precision_recall_curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEPCAYAAABY9lNGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x112758790>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "precision, recall, thresholds = precision_recall_curve(y_test_curve, lr.predict_proba(X_test_curve)[:, 1])\n",
    "plt.plot(recall, precision)\n",
    "plt.xlabel('recall')\n",
    "plt.ylabel('precision')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.93681718938102942"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import auc\n",
    "\n",
    "auc(recall, precision)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* ROC проходит через точки (0,0) и (1,1)\n",
    "* Если при каком-то значении порога $t$ алгоритм $a(x)$ идеально разделяет объекты 2 классов, то ROC проходит через точку (0,1).\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.metrics import roc_curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEPCAYAAABY9lNGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x112cb4490>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fpr, tpr, thresholds = roc_curve(y_test_curve, lr.predict_proba(X_test_curve)[:, 1])\n",
    "plt.plot(fpr, tpr)\n",
    "plt.xlabel('FPR')\n",
    "plt.ylabel('TPR')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9709965477077408"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "auc(fpr, tpr)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Валидация\n",
    "Чтобы оценить качество работы алгоритма, необходимо провести валидацию. Это один из самых важных шагов в процессе решения задачи. Оценим accuracy для модели логистической регрессии в задаче про спам-письма на тестовой выборке. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy = 0.814332247557\n"
     ]
    }
   ],
   "source": [
    "spam_data = pandas.read_csv(SPAMBASE_DATA_URL, header=None, names=(feature_names + ['spam']))\n",
    " \n",
    "X, y = spam_data.ix[:, :-1].values, spam_data.ix[:, -1].values\n",
    "\n",
    "# обучающая выборка\n",
    "X_tr = X[:n_train]\n",
    "y_tr = y[:n_train]\n",
    "\n",
    "# валидационная выборка\n",
    "X_test = X[n_train:]\n",
    "y_test = y[n_train:]\n",
    "\n",
    "# обучим ещё раз логистическую регрессию\n",
    "lr = LogisticRegression()\n",
    "lr = lr.fit(X_tr, y_tr)\n",
    "\n",
    "#посмотрим на точность классификации\n",
    "\n",
    "preds = lr.predict(X_test)\n",
    "print 'Accuracy =', metrics.accuracy_score(y_test, preds)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "А теперь попробуем перемешать объекты и повторим действия:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy = 0.914425427873\n"
     ]
    }
   ],
   "source": [
    "from sklearn import cross_validation\n",
    "X_tr, X_test, y_tr, y_test = cross_validation.train_test_split(X, y, test_size=train_part, random_state=123)\n",
    "\n",
    "# обучим ещё раз логистическую регрессию\n",
    "lr = LogisticRegression()\n",
    "lr = lr.fit(X_tr, y_tr)\n",
    "\n",
    "#посмотрим на точность классификации\n",
    "\n",
    "preds = lr.predict(X_test)\n",
    "print 'Accuracy =', metrics.accuracy_score(y_test, preds)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Как мы видим, качество классификации новых данных резко возросло. С чем это может быть связано? Рассмотрим вектор целевой переменной:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x1131d1810>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAEPCAYAAABRHfM8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nztrqMFIRA0wAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11326aa50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(y, 'ro')\n",
    "plt.xlabel('Object number')\n",
    "plt.ylabel('Target')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Проблема заключалась в том, что в выборке примеры были упорядочены: сначала шли примеры положительного класса, а потом отрицательного. Поэтому нельзя забывать **перемешивать классы**.\n",
    "\n",
    "Чтобы повысить устойчивость оценки качества, можно проводить разбиение выборки на обучающую и тестовую не один, $N$ раз, после чего усреднять результаты, полученные на $N$ контрольных выборках. Для этого можно использовать функцию [`sklearn.cross_validation.ShuffleSplit`](http://scikit-learn.org/stable/modules/generated/sklearn.cross_validation.ShuffleSplit.html#sklearn.cross_validation.ShuffleSplit)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "cv = cross_validation.ShuffleSplit(n=X.shape[0], n_iter = 5, train_size=0.9, test_size=0.1, random_state=123)\n",
    "quals = []\n",
    "lr = LogisticRegression()\n",
    "for tr_ind, test_ind in cv:\n",
    "    # using new set split for training and evaluating\n",
    "    lr.fit(X[tr_ind, :], y[tr_ind])\n",
    "    quals.append(metrics.roc_auc_score(y[test_ind], lr.predict_proba(X[test_ind,:])[:,1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean AUC-ROC = 0.972671869449\n",
      "AUC-ROC standart deviation = 0.00734595788166\n"
     ]
    }
   ],
   "source": [
    "print 'Mean AUC-ROC =', np.mean(quals)\n",
    "print 'AUC-ROC standart deviation =', np.std(quals)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Увеличим $N$:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "cv = cross_validation.ShuffleSplit(n=X.shape[0], n_iter = 10, train_size=0.9, test_size=0.1, random_state=124)\n",
    "quals = []\n",
    "lr = LogisticRegression()\n",
    "for tr_ind, test_ind in cv:\n",
    "    lr.fit(X[tr_ind, :], y[tr_ind])\n",
    "    quals.append(metrics.roc_auc_score(y[test_ind], lr.predict_proba(X[test_ind,:])[:,1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean AUC-ROC = 0.972414565272\n",
      "AUC-ROC standart deviation = 0.0037348645528\n"
     ]
    }
   ],
   "source": [
    "print 'Mean AUC-ROC =', np.mean(quals)\n",
    "print 'AUC-ROC standart deviation =', np.std(quals)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Также рассмотрим зависимость оценки качества от пропорции, в которой выборка разбивается на обучающую и тестовую:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "means = []\n",
    "stds = []\n",
    "xs = []\n",
    "quals = []\n",
    "for tr_size in np.arange(0.5,0.9,0.01):\n",
    "    # set train size\n",
    "    cv = cross_validation.ShuffleSplit(n=X.shape[0], n_iter = 5, train_size=tr_size, random_state=123)\n",
    "    fold_quals = []\n",
    "    lr = LogisticRegression()\n",
    "    for tr_ind, test_ind in cv:\n",
    "        # evaluate quality for each split\n",
    "        lr.fit(X[tr_ind, :], y[tr_ind])\n",
    "        qual = metrics.roc_auc_score(y[test_ind], lr.predict_proba(X[test_ind,:])[:,1])\n",
    "        fold_quals.append(qual)\n",
    "        quals.append(qual)\n",
    "        xs.append(tr_size)\n",
    "    # evaluation for current train set size\n",
    "    means.append(np.mean(fold_quals))\n",
    "    stds.append(np.std(fold_quals))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x10b794d50>"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZsAAAEPCAYAAACUb2mtAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nZtn9f6j0S3RpkDqOAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113315790>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "figsize(6,4)\n",
    "plt.plot(np.arange(0.5,0.9,0.01), means, 'ro')\n",
    "plt.xlabel('train size')\n",
    "plt.ylabel('AUC-ROC')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x112ca1850>"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZsAAAEKCAYAAADEovgeAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nbEREJHb/D8gUrZtPYHnpAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x110ce4ed0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(np.arange(0.5,0.9,0.01), stds)\n",
    "plt.xlabel('train size')\n",
    "plt.ylabel('AUC-ROC std')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Несбалансированные классы\n",
    "\n",
    "Если объём выборки невелик, а объектов одного класса значительно меньше, чем другого, то может сложиться ситуация, когда при случайном разбиении объектов меньшего класса не окажется в тестовой выборке, в связи с чем результаты оценки качества будут неустойчивы."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Пример**: задача кредитного скоринга. Поскольку случаи невозврата кредита довольно редки, количество объектов отрицательного класса будет значительно меньше, чем положительного."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Доля заемщиков, не вернувших кредит: 0.00712606000143\n"
     ]
    }
   ],
   "source": [
    "df = pd.read_csv('data/givemesomecredit')\n",
    "X = df.drop('SeriousDlqin2yrs', axis=1)\n",
    "X = X.fillna(X.mean()).as_matrix()\n",
    "y = df['SeriousDlqin2yrs']\n",
    "print \"Доля заемщиков, не вернувших кредит:\", y.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Всего 0.7% выборки составляют объекты положительного класса"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- В таком случае необходимо производить стратификацию, то есть разбивать отдельно объекты каждого класса на обучение и тест (сохраняя их доли).\n",
    "\n",
    "Оценим влияние стратификации на оценку качества путем разбиения выборки $N=10$ раз на обучение и тест и последующего усреднения AUC-ROC на тестовой выборке:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AUC-ROC w/o stratification =  0.644737388024\n",
      "AUC-ROC std w/o stratification =  0.0784631563488\n"
     ]
    }
   ],
   "source": [
    "cv = cross_validation.ShuffleSplit(n=X.shape[0], n_iter = 10, train_size=0.8, random_state=124)\n",
    "quals = []\n",
    "\n",
    "lr = LogisticRegression()\n",
    "for tr_ind, test_ind in cv:\n",
    "    lr.fit(X[tr_ind, :], y[tr_ind])\n",
    "    quals.append(metrics.roc_auc_score(y[test_ind], lr.predict_proba(X[test_ind,:])[:,1]))\n",
    "\n",
    "print \"AUC-ROC w/o stratification = \", np.mean(quals)\n",
    "print \"AUC-ROC std w/o stratification = \", np.std(quals)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AUC-ROC with stratification =  0.626262553802\n",
      "AUC-ROC std with stratification =  0.0590682738761\n"
     ]
    }
   ],
   "source": [
    "cv = cross_validation.StratifiedShuffleSplit(y, n_iter = 10, train_size=0.8, random_state=124)\n",
    "quals = []\n",
    "lr = LogisticRegression()\n",
    "for tr_ind, test_ind in cv:\n",
    "    lr.fit(X[tr_ind, :], y[tr_ind])\n",
    "    quals.append(metrics.roc_auc_score(y[test_ind], lr.predict_proba(X[test_ind,:])[:,1]))\n",
    "\n",
    "print \"AUC-ROC with stratification = \", np.mean(quals)\n",
    "print \"AUC-ROC std with stratification = \", np.std(quals)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Как мы видим, стратификация позволяет уменьшить дисперсию и более точно оценить качество"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Кросс-валидация\n",
    "Рассмотрим датасет о пациентах, больных бесплодием, содержащий около 100 объектов. Выборка небольшая, но из специфичной области, поэтому каждый объект может нести в себе важную информацию, влияющую на значение целевой переменной. В связи с этим при оценивании качества модели хотелось бы сделать обучающую выборку как можно больше. При этом из-за маленького объема всей выборки мы можем позволить себе обучать модель многократно. В данной ситуации для оценки качества можно использовать схему leave-one-out, реализованную в качестве объекта [`LeaveOneOut`](http://scikit-learn.org/stable/modules/generated/sklearn.cross_validation.LeaveOneOut.html#sklearn.cross_validation.LeaveOneOut).\n",
    "\n",
    "Оценка качества производится следующим образом: каждый объект поочередно исключается из обучающей выборки, после чего модель обучается на всех остальных объектах, а качество измеряется на исключенном объекте (этот объект играет роль тестовой выборки); после того, как каждый объект был выкинут ровно по одному разу, итоговая оценка качества получается как среднее по всем полученным значениям функционала:\n",
    "\n",
    "$$Q_{LOO}(X) = \\frac{1}{l} \\sum_{i=1}^l Q(\\{x_i \\}; a_{X \\backslash \\{x_i\\} }),$$\n",
    "где\n",
    "* $X = \\left\\{ (x_i, y_i)\\right\\}_{i=1}^l$ — обучающая выборка;\n",
    "* $Q(X^{test}; a)$ — значение функционала алгоритма $a$ на контрольной выборке $X^{test}$;\n",
    "* $a_{X^{train}}(\\cdot)$ — алгоритм, обученный на обучающей выборке $X^{train}.$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LOO accuracy = 0.88\n"
     ]
    }
   ],
   "source": [
    "df = pd.read_csv('data/fertility', sep = ',', header = None)\n",
    "X = df.ix[:,:8]\n",
    "y = np.zeros(df.shape[0])\n",
    "y[df[9].as_matrix() == 'N'] = 1\n",
    "\n",
    "cv_LOO = cross_validation.LeaveOneOut(n = len(y))\n",
    "qual = []\n",
    "for tr_ids, test_ids in cv_LOO:\n",
    "    lr.fit(X.ix[tr_ids,:], y[tr_ids])\n",
    "    qual.append(lr.predict(X.ix[test_ids, :]) == y[test_ids])\n",
    "print \"LOO accuracy =\", np.mean(qual)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "С ростом размера выборки возрастают также и затраты на многократное обучение модели. Тем не менее, хотелось бы, чтобы каждый объект побывал и в обучающей, и в тестовой выборках, причём одинаковое количество раз. Чтобы удовлетворить этому условию, можно использовать схему K-fold кросс-валидации, реализованную в качестве объекта [`sklearn.cross_validation.KFold`](http://scikit-learn.org/stable/modules/generated/sklearn.cross_validation.KFold.html#sklearn.cross_validation.KFold).\n",
    "\n",
    "![](http://scott.fortmann-roe.com/docs/docs/MeasuringError/crossvalidation.png)\n",
    "\n",
    "В данном случае выборка разбивается на $K$ (примерно) одинаковых блоков, каждый из которых поочередно выступает в качестве контрольной выборки, а оставшиеся $K-1$ — в качестве обучающей, после чего оценки качества усредняются:\n",
    "\n",
    "$$Q_{K-fold}(X) = \\frac{1}{K} \\sum_{k=1}^K Q(X^k; a_{X \\backslash X^k }),$$\n",
    "где $X^k, k = \\overline{1, K}$ — непересекающиеся блоки, на которые разбивается выборка $X$: $X = \\sqcup_{k=1}^K X^k.$\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "K-fold accuracy = 0.88\n"
     ]
    }
   ],
   "source": [
    "cv = cross_validation.KFold(n=len(y), n_folds=5, random_state=123, shuffle=True)\n",
    "lr = LogisticRegression()\n",
    "quals = []\n",
    "for tr_ids, test_ids in cv:\n",
    "    lr.fit(X.ix[tr_ids,:], y[tr_ids])\n",
    "    quals.append(metrics.accuracy_score(y[test_ids], lr.predict(X.ix[test_ids, :])))\n",
    "\n",
    "print \"K-fold accuracy =\", np.mean(quals)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Вопросы:\n",
    "1. Как соотносятся кросс-валидация по схеме K-fold и схема leave-one-out?\n",
    "2. Что общего у k-fold и ShuffleSplit?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Настройка параметров модели\n",
    "\n",
    "Настройка наилучших параметров модели производится с использованием техник кросс-валидации (для оценки качества) и соответственно перебора параметров. \n",
    "\n",
    "Для перебора параметров можно использовать: \n",
    "- Модуль для перебора параметров модели по сетке [`sklearn.grid_search`](http://scikit-learn.org/stable/modules/classes.html#module-sklearn.grid_search)\n",
    "  - Класс [`GridSearchCV`](http://scikit-learn.org/stable/modules/generated/sklearn.grid_search.GridSearchCV.html#sklearn.grid_search.GridSearchCV) — полный перебор параметров модели по сетке\n",
    "  - Класс [`RandomizedSearchCV`](http://scikit-learn.org/stable/modules/generated/sklearn.grid_search.RandomizedSearchCV.html#sklearn.grid_search.RandomizedSearchCV) — перебирает случайные комбинации параметров"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Краткий справочник по SciKit-Learn\n",
    "\n",
    "Обратите внимание на следующую функциональность библиотеки SciKit-Learn:\n",
    "\n",
    "- Модуль для оценки качества моделей и кросс валидации [`sklearn.cross_validation`](http://scikit-learn.org/stable/modules/cross_validation.html):\n",
    "  - функция разбиения выборки [`train_test_split()`](http://scikit-learn.org/stable/modules/generated/sklearn.cross_validation.train_test_split.html#sklearn.cross_validation.train_test_split)\n",
    "  - итераторы разбиений для кросс-валидации: [`LeaveOneOut`](http://scikit-learn.org/stable/modules/generated/sklearn.cross_validation.LeaveOneOut.html#sklearn.cross_validation.LeaveOneOut), [`KFold`](http://scikit-learn.org/stable/modules/generated/sklearn.cross_validation.KFold.html#sklearn.cross_validation.KFold), [`ShuffleSplit`](http://scikit-learn.org/stable/modules/generated/sklearn.cross_validation.ShuffleSplit.html#sklearn.cross_validation.ShuffleSplit)\n",
    "  - всегда указывайте параметр `random_state`, это сделает код вашего эксперимента воспроизводимым: при повторном запуске разбиения будут одинаковыми\n",
    "- Модуль с метриками качества моделей [`sklearn.metrics`](http://scikit-learn.org/stable/modules/classes.html#module-sklearn.metrics)\n",
    "  - [`accuracy_score(y_true, y_pred)`](http://scikit-learn.org/stable/modules/generated/sklearn.metrics.accuracy_score.html#sklearn.metrics.accuracy_score)\n",
    "  - [`precision_recall_fscore_support(y_true, y_pred)`](http://scikit-learn.org/stable/modules/generated/sklearn.metrics.precision_recall_fscore_support.html#sklearn.metrics.precision_recall_fscore_support)\n",
    "  - [`roc_curve(y_true, y_score)`](http://scikit-learn.org/stable/modules/generated/sklearn.metrics.roc_curve.html#sklearn.metrics.roc_curve)\n",
    "  - [`precision_recall_curve(y_true, y_score)`](http://scikit-learn.org/stable/modules/generated/sklearn.metrics.precision_recall_curve.html)\n",
    "  - [`roc_auc_score(y_true, y_score)`](http://scikit-learn.org/stable/modules/generated/sklearn.metrics.roc_auc_score.html#sklearn.metrics.roc_auc_score)\n",
    "\n",
    "- Модуль для перебора параметров модели по сетке [`sklearn.grid_search`](http://scikit-learn.org/stable/modules/classes.html#module-sklearn.grid_search)\n",
    "  - Класс [`GridSearchCV`](http://scikit-learn.org/stable/modules/generated/sklearn.grid_search.GridSearchCV.html#sklearn.grid_search.GridSearchCV) — полный перебор параметров модели по сетке\n",
    "  - Класс [`RandomizedSearchCV`](http://scikit-learn.org/stable/modules/generated/sklearn.grid_search.RandomizedSearchCV.html#sklearn.grid_search.RandomizedSearchCV) — перебирает случайные комбинации параметров\n",
    "  \n",
    "### Руководства и примеры из документации\n",
    "\n",
    "- [Оценка качества предсказаний](http://scikit-learn.org/stable/modules/model_evaluation.html)\n",
    "- [Оптимизация параметров по сетке с использованием кросс-валидации](http://scikit-learn.org/stable/auto_examples/grid_search_digits.html)\n",
    "- [ROC-кривые](http://scikit-learn.org/stable/auto_examples/plot_roc.html), [Precision-Recall кривые](http://scikit-learn.org/stable/auto_examples/plot_precision_recall.html)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}