{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 17 - Predictive Models 101\n",
    "\n",
    "이 책의 1부에서는 인과 추론의 핵심 내용을 배웠습니다. 1부에서 소개한 개념은 잘 정의되었으며 오랫동안 검증되어온 것입니다. 앞으로 여정의 단단한 기반이 될 것입니다. 지금까지 인과 추론, 편향의 의미를 배웠습니다. 또한, 편향을 다루는 여러 방법(선형 회귀, `matching`, 성향 점수, 도구 변수, `Diff-in-Diff`, `RDD` 등)을 공부했습니다. 1부에서는 주로 평균 처치 효과 $E[Y_1 - Y_0]$를 구하기 위한 기초적인 기술에 초점을 맞췄습니다.\n",
    " \n",
    "2부에서는 기계학습을 활용한 최근의 인과추론 연구 결과와 응용 방법을 배울 예정입니다. 이론적인 엄격함보다는 실용적으로 접근합니다. 일부 기술은 견고한 이론이 없지만 경험적으로 효과가 입증되었습니다. 근본적인 인과 관계를 연구하려는 과학자들보다는 실무에 인과 추론을 적용하고자 하는 사람에게 더 유용할 것입니다.\n",
    " \n",
    "먼저 `heterogeneous treatment effects`(이종 처치 효과)의 개념에 관해 공부할 예정입니다. 평균 처치 효과인 $E[Y_1 - Y_0]$가 아닌 $X$를 조건으로 하는 처치 효과 $E[Y_1 - Y_0 | X]$를 구하고자 합니다. 요즘은 개인화가 중요한 세상입니다. 효과 있는 특정 그룹에 집중하고자 합니다. 이러한 상황에서 다음과 같은 질문에 답하고자 합니다. 우리는 누구를 신경 써야 할까요?\n",
    "\n",
    "이 문제는 대부분 회사의 고민이기도 합니다. 누구에게 할인권을 줄까요? 대출 이자는 얼마가 적당할까요? 고객에게 어떤 물건을 보여줄까요? 사이트에 접속할 때 어떤 페이지를 띄워줘야 할까요? 이는 모두 이종 처치 효과에 관한 질문입니다. 더 깊이 들어가기 전에 앞으로 기본 도구가 될 기계학습이 산업에 미친 영향을 살펴봅시다.\n",
    "\n",
    "## Machine Learning in the Industry\n",
    " \n",
    "산업적으로 **기계학습**을 어떻게 쓰이고 있을까요? 17장은 기계학습에 익숙하지 않은 사람에게 좋은 단기 집중 과정이 될 것입니다. 기계학습을 처음 접하는 사람은 앞으로 배울 기술을 최대한 활용하기 위해 기초 내용을 숙지하는 것을 강력히 추천해 드립니다. 기계학습에 관해 잘 알고 있더라도 읽어주시기를 바랍니다. 끝까지 읽으면 분명 얻는 게 있을 거예요. 여기서는 일반적인 기계학습 설명 자료와는 달리 의사 결정 나무나 신경망 등 알고리즘을 자세히 설명하지는 **않을** 예정입니다. 대신 **실제 세계에서 기계학습을 어떻게 사용하는지** 알아보겠습니다.\n",
    " \n",
    "![img](./data/img/industry-ml/ml-meme.png)\n",
    " \n",
    "인과추론에서 기계학습이 중요한 이유는 무엇일까요? 인과관계는 기계학습의 예측과 대비됩니다. 기계학습과 비교함으로써 인과추론을 더 잘 이해할 수 있습니다. 또한, 최근의 인과추론은 기계학습 알고리즘을 활용하려는 시도가 많습니다. 이러한 내용을 이해하기 위해서는 기계학습을 잘 알고 있어야 합니다. 다만 기계학습은 마법의 기술처럼 과장된 면이 있어 실제로 기계학습의 작동 방법을 설명할 필요가 있습니다.\n",
    "\n",
    "**기계학습은 빠르게 자동으로 예측을 수행하는 방법입니다.** 이 문장만으로도 기계학습의 90%는 설명할 수 있습니다. 컴퓨터 비전, 자율주행, 언어번역, 질병 진단과 같은 분야의 발전은 지도 기계학습의 발전에 따라 이루어졌습니다. 이러한 일들이 일이 예측으로 보이지 않을 수 있습니다. 어떻게 언어번역이 예측 문제인가요? 모든 문제를 예측 문제로 만드는 것은 기계학습의 특징입니다. 생각보다 많은 문제를 예측 문제로 바꿔 해결할 수 있습니다. 언어번역은 문장을 모델에 제시하고 다른 언어의 문장을 예측하는 것으로 해결할 수 있습니다. 여기서 예측은 **미래에 대한 예측이 아닙니다.** 대신 입력값으로부터 잘 정의된 출력값으로 대응(`mapping`)시키는 것을 의미합니다.\n",
    " \n",
    "![img](./data/img/industry-ml/translation.png)\n",
    " \n",
    "기계학습 모델이 하는 일은 매우 복잡한 `mapping` 함수를 학습하는 것입니다. 풀고자 하는 문제를 입력값과 출력값을 대응시키는 문제로 바꿀 수 있다면 기계학습은 좋은 접근입니다. 자율주행 또한 출력값이 여럿인 예측 문제로 생각할 수 있습니다. 바퀴 각도, 브레이크 압력, 가속기 압력을 예측하는 문제를 해결함으로써 자율주행 자동차를 구현할 수 있습니다.\n",
    "\n",
    "수학적으로 기계학습은 다음과 같은 기댓값을 추정합니다.\n",
    " \n",
    "$\n",
    "E[Y|X]\n",
    "$\n",
    " \n",
    "$Y$는 우리가 알고 싶은 것(번역된 문장, 질병 진단 결과)이며, $X$는 이미 알고 있는 것(입력 문장, X선 사진)입니다. 기계학습은 단순히 조건부 기대 함수를 추정합니다.\n",
    "\n",
    "좋습니다. 처음 생각보다 기계학습의 예측은 훨씬 강력해 보입니다. 자율주행 자동차나 언어번역 문제를 해결하는 것은 꽤 멋진 일이지만, 구글이나 우버에서 일하지 않는 한 여러분의 일과는 거리가 있을 겁니다. 좀 더 일반적인 예시로 고객을 확보하는 문제를 생각해보겠습니다.\n",
    " \n",
    "고객을 확보하기 전에는 어떤 고객이 수익을 가져다주는지 파악해야 합니다. 고객을 확보하기 위해서는 비용(광고비, 배송비 등)이 필요하며 확보된 고객은 여러분 회사에 수익을 가져다줍니다. 인터넷 사업자나 가스 공급 회사를 생각하면 이해하기가 쉽습니다. 일반적으로 고객 별 현금 흐름은 다음과 같습니다.\n",
    " \n",
    "![img](./data/img/industry-ml/cashflow-1.png)\n",
    " \n",
    "그림에서 막대기는 비용이나 수익과 같은 금전적인 사건을 의미합니다. 고객을 확보하기 위해 광고비를 투자(빨간색)해야 합니다. 고객을 확보하고 나서도 온보딩(제품 사용 설명)이나 설치 비용이 발생할 수 있습니다. 모든 투자가 끝난 후에야 수익이 발생하기 시작합니다. 특정 시점에서는 유지보수 비용이 발생할 수 있습니다. 최종적으로 고객과의 계약이 종료되면 수익과 비용을 결산합니다.\n",
    " \n",
    "어떤 고객이 수익을 가져다주는지 알기 위해서 막대기를 계단식으로 정렬할 수 있습니다. 비용과 수익의 합계가 $0$보다 클수록 수익이 많습니다.\n",
    "\n",
    "![img](./data/img/industry-ml/cascade-1.png)\n",
    " \n",
    "반대로 비용이 더 클 수도 있습니다. 유지보수 비용이 크다면 계단식 그래프에서 결과가 $0$ 아래로 떨어질 수 있습니다.\n",
    " \n",
    "![img](./data/img/industry-ml/cascade-2.png)\n",
    " \n",
    "실제 현금 흐름은 사업에 따라 더 복잡할 수 있습니다. 이자율이나 할인에 관한 내용을 추가할 수도 있지만 설명에 필요한 내용은 이것으로 충분합니다.\n",
    "\n",
    "우리는 무엇을 할 수 있을까요? 고객 데이터를 많이 가지고 있다면 수익에 따라 고객을 구분하기 위한 기계학습 모델을 훈련할 수 있습니다. 학습된 모델의 예측값으로 수익이 많이 나는 고객에 집중할 수 있습니다. 또한, 고객이 큰 비용을 발생시키기 전 계약을 해지할 수도 있습니다. 우리는 **문제를 예측 문제로 만들어 기계학습으로 풀고자 합니다.** 즉, 수익이 나는 고객을 식별하여 이들과 거래하고 싶습니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "tags": [
     "hide-input"
    ]
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn import ensemble\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import r2_score\n",
    "import seaborn as sns\n",
    "from matplotlib import pyplot as plt\n",
    "from matplotlib import style\n",
    "style.use(\"ggplot\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "여기 $10,000$명의 고객에 관한 30일간 거래 데이터가 있습니다. `cacq`는 고객별로 발생하는 비용을 의미합니다. 배송비나 제품 교육비가 이에 해당합니다. 여기서는 고객을 완벽하게 통제할 수 있다고 가정합니다. 당신은 특정 고객과 거래하거나 거절할 수 있습니다. 사전에 수익을 낼 수 있는 고객을 파악하고 그들하고만 거래하려 합니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10000, 32)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>customer_id</th>\n",
       "      <th>cacq</th>\n",
       "      <th>day_0</th>\n",
       "      <th>day_1</th>\n",
       "      <th>day_2</th>\n",
       "      <th>day_3</th>\n",
       "      <th>day_4</th>\n",
       "      <th>day_5</th>\n",
       "      <th>day_6</th>\n",
       "      <th>day_7</th>\n",
       "      <th>...</th>\n",
       "      <th>day_20</th>\n",
       "      <th>day_21</th>\n",
       "      <th>day_22</th>\n",
       "      <th>day_23</th>\n",
       "      <th>day_24</th>\n",
       "      <th>day_25</th>\n",
       "      <th>day_26</th>\n",
       "      <th>day_27</th>\n",
       "      <th>day_28</th>\n",
       "      <th>day_29</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>-110</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>73</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>21</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>-58</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>15</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>-7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>-30</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>40</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>-42</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 32 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   customer_id  cacq  day_0  day_1  day_2  day_3  day_4  day_5  day_6  day_7  \\\n",
       "0            0  -110      6      0     73     10      0      0      0     21   \n",
       "1            1   -58      0      0      0     15      0      3      2      0   \n",
       "2            2    -7      0      0      0      0      0      0      0      1   \n",
       "3            3   -30      0      3      2      0      9      0      0      0   \n",
       "4            4   -42      0      0      0      0      0      0      0      0   \n",
       "\n",
       "   ...  day_20  day_21  day_22  day_23  day_24  day_25  day_26  day_27  \\\n",
       "0  ...       0       0       0       0       0       0       0       0   \n",
       "1  ...       0       0       0       0       0       0       0       0   \n",
       "2  ...       0       0       0       0       0       0       0       0   \n",
       "3  ...       0       0      40       0       0       0       0       0   \n",
       "4  ...       0       0       0       0       0       0       0       0   \n",
       "\n",
       "   day_28  day_29  \n",
       "0       0       0  \n",
       "1       0       0  \n",
       "2       0       0  \n",
       "3       0       0  \n",
       "4       0       0  \n",
       "\n",
       "[5 rows x 32 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transactions = pd.read_csv(\"data/customer_transactions.csv\")\n",
    "print(transactions.shape)\n",
    "transactions.head()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "기계학습은 거래 데이터로부터 좋은 고객과 나쁜 고객을 구분하는 규칙을 학습할 것입니다. 문제를 단순히 하기 위해 모든 거래 데이터와 `cacq`를 더하겠습니다. 이 방법은 거래 내역을 감춘다는 사실에 유의하세요.\n",
    " \n",
    "합계는 `net_value`로 표기합니다. 우리의 목표는 고객과 계약을 결정하기 **전** 수익이 발생할 고객을 파악하는 것이므로 미리 알 수 있는 데이터만 `feature`로 사용합니다. `feature`에 해당하는 변수는 지역(`region`), 소득(`income`), 나이(`age`)로 별도의 `csv`파일에서 가져옵니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>customer_id</th>\n",
       "      <th>region</th>\n",
       "      <th>income</th>\n",
       "      <th>age</th>\n",
       "      <th>net_value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>30</td>\n",
       "      <td>1025</td>\n",
       "      <td>24</td>\n",
       "      <td>130</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>41</td>\n",
       "      <td>1649</td>\n",
       "      <td>26</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>18</td>\n",
       "      <td>2034</td>\n",
       "      <td>33</td>\n",
       "      <td>-6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>20</td>\n",
       "      <td>1859</td>\n",
       "      <td>35</td>\n",
       "      <td>136</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1243</td>\n",
       "      <td>26</td>\n",
       "      <td>-8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   customer_id  region  income  age  net_value\n",
       "0            0      30    1025   24        130\n",
       "1            1      41    1649   26         10\n",
       "2            2      18    2034   33         -6\n",
       "3            3      20    1859   35        136\n",
       "4            4       1    1243   26         -8"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "profitable = (transactions[[\"customer_id\"]]\n",
    "              .assign(net_value = transactions\n",
    "                      .drop(columns=\"customer_id\")\n",
    "                      .sum(axis=1)))\n",
    "\n",
    "customer_features = (pd.read_csv(\"data/customer_features.csv\")\n",
    "                     .merge(profitable, on=\"customer_id\"))\n",
    "\n",
    "customer_features.head()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "좋습니다! 조금씩 문제를 정리하고 있습니다. 기계학습 모델로 돈이 되는 고객 (`net_value > 0`)을 파악하고자 합니다. 여러 방법을 시도해보고 효과적인 방법을 알아봅시다. 하지만 그 전에 기계학습이 작동하는 방법을 간단히 살펴봅시다. 이미 기계학습이 어떻게 작동하는지 알고 있다면 건너뛰어도 무방합니다.\n",
    " \n",
    "## Machine Learning Crash Course\n",
    " \n",
    "기계학습이 작동하기 위해서는 정답지(`label`)가 필요합니다. 수집한 데이터를 기계학습 모델로 훈련하면 새로운 데이터를 입력으로 받아 예측값을 얻을 수 있습니다. 아래 그림은 전체적인 흐름을 나타냅니다.\n",
    " \n",
    "![img](./data/img/industry-ml/ml-flow.png)\n",
    " \n",
    "예시에서 `label`은 `net_value`입니다. `feature`(`region`, `income`, `age`)로 `net_value`를 예측하는 기계학습 모델을 훈련하면 새로운 데이터로 `net_value`를 예측할 수 있습니다. 예측은 그림의 왼쪽 부분에 해당합니다. 새로운 데이터는 `feature`(`region`, `income`, `age`)만 알고 있지만 기계학습 모델을 사용해 `net_value`를 예측할 수 있습니다.\n",
    "\n",
    "기계학습은 조건부 기댓값 $E[Y|X]$을 추정합니다. $Y$는 `target variable` 이나 `outcome`, $X$는 `feature`라 불립니다. 기계학습은 단지 $\\hat{E}[Y|X]$를 얻을 방법의 하나일 뿐입니다. 일반적으로는 오차나 손실 함수를 최적화하는 방법으로 학습할 수 있습니다.\n",
    " \n",
    "기계학습은 어떠한 함수라도 근사할 수 있어 까다롭습니다. 항상 훈련 데이터를 완벽하게 맞출 수 있을 정도로 모델을 강력하게 만들 수 있습니다. 대부분 기계학습 모델은 복잡성을 조절하는 하이퍼 파라미터가 있으며 모델의 복잡도를 조정할 수 있습니다. 아래 그림은 단순한 모델(왼쪽), 중간 모델(가운데) 및 복잡한 모델(오른쪽)을 보여줍니다. 복잡한 모델은 훈련 데이터를 완벽하게 맞추고 있습니다.\n",
    "\n",
    "![img](./data/img/industry-ml/model-fit.png)\n",
    " \n",
    "모델을 배포하기 전에 미리 평가할 방법이 있을까요? 단순하게 예측값과 실제값을 비교해 볼 수 있습니다. $R^2$와 같은 지표를 사용하면 비교 결과를 수치로 확인할 수 있습니다. 복잡한 모델은 데이터를 완벽하게 맞추고 있으므로 예측값은 실제값과 완벽히 일치합니다. 이 방법은 모델을 복잡하게 만들수록 결과가 좋아진다는 점에서 문제가 있습니다.\n",
    "\n",
    "매우 복잡한 모델은 일반적으로 **좋지 않습니다**. 여러분은 어떤 모델을 선호하시나요? 복잡한 모델이요? 아마도 가운데 모델을 선택할 것입니다. 데이터에 완벽하게 맞지는 않지만, 여전히 좋은 예측을 할 수 있기 때문입니다.\n",
    " \n",
    "![img](./data/img/industry-ml/overfitting.jpg)\n",
    " \n",
    "모델을 너무 복잡하게 만들면 데이터의 패턴뿐 아니라 무작위 잡음도 함께 학습합니다. 학습된 잡음은 실제 잡음과 다르므로 결국 \"완벽한\" 모델은 틀리게 됩니다. 전문용어로는 모델이 과적합 되어 있어 일반화가 어렵다고 합니다. 어떻게 하면 좋을까요?\n",
    "\n",
    "아이디어는 정답을 알고 있는 데이터를 두 개로 나누는 것입니다. 하나는 모델 훈련에 사용하고, 다른 하나는 모델 검증에 사용합니다. 두 번째 데이터는 훈련 중 접근할 수 없는 것처럼 생각합니다. 이 방법은 교차검증이라 부릅니다. \n",
    "\n",
    "![img](./data/img/industry-ml/test.png)\n",
    " \n",
    "위 그림은 검증 결과 보여줍니다. 복잡한 모델은 잘 맞지 않습니다. 오히려 가운데 모델이 더 좋아 보입니다. 교차검증은 매우 중요한 내용이므로 더 살펴보겠습니다."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Cross Validation\n",
    " \n",
    "교차검증은 모델 선택에 유용한 방법입니다. 거의 필수적으로 채용하고 있습니다. 교차검증의 기본 아이디어는 실제 세계를 모방하는 것입니다. 이미 정답을 알고 있지만 모델이 보지 않은 데이터(`unseen data`)를 예측해 봅니다. `unseen data`는 앞으로 마주할 새로운 데이터를 대신합니다.\n",
    " \n",
    "고객 확보 문제에 교차검증을 적용해보겠습니다. 아래는 순서에 따라 진행합니다.\n",
    "\n",
    "1. 훈련 데이터와 검증 데이터를 나눕니다. 훈련 데이터는 이미 정답을 알고 있으므로 기계학습 모델을 학습할 수 있습니다. \n",
    "2. 훈련 데이터로 임의의 고객으로부터 얻을 수익을 예측하는 규칙을 **학습**합니다.\n",
    "3. **학습에 사용하지 않은** 데이터에 규칙을 적용해 봅니다. 즉, 훈련 데이터로 학습된 모델을 검증 데이터에 적용합니다. \n",
    " \n",
    "아래 그림은 분할된 데이터를 나타냅니다. 그림의 가장 오른쪽에는 정말로 볼 수 없는 데이터(`Unseen Data`)가 있으며 내부 데이터(`Internal Data`) 안에서 보이지 않는 것 가정하는 검증 데이터(`Test Data`)를 얻습니다.\n",
    "\n",
    "![img](./data/img/industry-ml/cross-validation.png)\n",
    " \n",
    "쉽게 말하자면 내부 데이터로 훈련 데이터와 테스트 데이터로 나눕니다. 훈련 데이터로 수익을 예측하는 규칙을 학습하고, 테스트 데이터로 검증합니다. 테스트 데이터는 학습하지 않습니다.\n",
    "\n",
    "교차검증 방법으로 `k-fold cross-validation`, `temporal cross validation` 등 더 복잡한 방법이 있지만 지금은 단순하게 둘로 나누는 정도면 충분합니다. 교차검증의 기본 아이디어는 배포 환경에서 어떤 일이 생길지 시뮬레이션하여 현실적인 추정치를 얻는 것입니다.\n",
    "\n",
    "데이터를 둘로 나눠 보겠습니다. $70$%는 모델 학습에 사용하며 나머지 $30$%는 테스트에 사용합니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((7000, 5), (3000, 5))"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train, test = train_test_split(customer_features, test_size=0.3, random_state=13)\n",
    "train.shape, test.shape"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Predictions and Policies\n",
    " \n",
    "![img](./data/img/industry-ml/profit.png)\n",
    " \n",
    "두 가지 개념을 상세히 살펴보겠습니다. **prediction**(예측)은 무언가 추정하거나 예측하는 숫자로 $\\hat{E}[y_i|X_i]$와 같습니다. 앞의 예시에서는 예측된 수익이며, 예측 결과는 `16BRL`와 같이 단순한 숫자로 표현됩니다.\n",
    " \n",
    "두 번째 개념은 **policy**(정책)입니다. 정책은 결정을 자동화하는 규칙을 뜻합니다. 예를 들면 예상 수익이 $1,000$ 이상인 고객과만 계약하는 규칙을 만들 수 있습니다. 일반적으로 예측을 기반으로 정책을 만듭니다. 수익이 $10$ 이하인 고객을 계약에서 제외하고자 한다면 정책은 $\\hat{E}[y_i|X_i] > 10$가 됩니다. 기계학습으로는 첫 번째 개념인 예측을 잘 다룰 수 있습니다. 하지만 예측만으로는 할 수 있는 것은 없습니다. 예측 결과를 기반으로 정책을 만들어 결정해야 합니다.\n",
    "\n",
    "예측과 정책 모두 교차검증을 적용해야 합니다. 즉, 하나의 데이터 세트로 예측 모델과 정책을 만들고 다른 데이터 세트로 확인해야 합니다.\n",
    " \n",
    "## One Feature Policies\n",
    " \n",
    "기계학습으로 넘어가기 전에 간단한 것들을 시도해 보겠습니다. $20$%만 노력해도 $80$%의 성과를 낼 수 있지만 종종 대부분의 데이터 과학자는 이 사실을 잊습니다. 우리가 할 수 있는 가장 간단한 일은 무엇일까요? **모든 고객과 계약합니다!** 누가 수익을 가져다주는지 파악하는 대신 모든 고객과 계약하고 총수입이 많기를 기대하는 것입니다.\n",
    " \n",
    "모든 고객과 계약하는 것이 좋은 생각인지 평균 순수익을 확인해 봅시다. 순수익이 양수라면 평균적으로 수익을 낸다는 의미입니다. 고객별로 수익이 날 수도 손해 볼 수도 있지만, 고객을 충분히 많이 확보한다면 결국 돈을 벌 수 있을 것입니다. 반면에 순수익이 음수면 모든 고객과 계약할 때 손해 보게 됩니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-29.169428571428572"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train[\"net_value\"].mean()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "실망스럽게도 모든 고객과 계약하면 평균적으로 $30$헤알만큼 손해 봅니다. 아주 단순한 정책은 효과가 없습니다. 수익을 내려면 더 좋은 방법을 찾아야 합니다. 참고로 이 예시는 교육적인 목적이 큽니다. 단순한 정책인 \"모든 사람을 똑같이 대한다.\"는 효과가 없었지만, 실제로는 효과가 좋을 때가 더 많습니다. 일반적으로 모든 사람에게 광고 메일을 보내는 것이 좋으며, 모든 사람에게 할인 쿠폰을 배포하는 것도 좋습니다.\n",
    " \n",
    "그다음으로 간단한 방법은 무엇일까요? 고객의 `feature`로 좋은 고객과 나쁜 고객을 구별해 봅시다. 고객의 소득을 예로 들면 직관적으로 부유한 고객이 더 많은 수익을 낼 것으로 생각할 수 있습니다. 부유한 고객과만 계약한다면 어떨까요? 좋은 생각인지 확인해 봅시다.\n",
    " \n",
    "분석을 위해 데이터를 소득 분위수로 나눕니다. 분위수는 데이터를 동일한 크기로 나눌 수 있어 개인적으로 선호하는 방법입니다. 그리고 각각의 소득 분위수의 평균 순이익을 계산해 보겠습니다. 기댓값은 음의 평균 순이익 $E[NetValue]<0$이지만, 순이익이 양수인 $E[NetValue|Income=x]<0$ (아마도 더 높은 소득 그룹) 일부 하위 모집단이 있을 수 있습니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,6))\n",
    "np.random.seed(123) ## seed because the CIs from seaborn uses boostrap\n",
    "\n",
    "# pd.qcut create quantiles of a column\n",
    "sns.barplot(data=train.assign(income_quantile=pd.qcut(train[\"income\"], q=20)), \n",
    "            x=\"income_quantile\", y=\"net_value\")\n",
    "plt.title(\"Profitability by Income\")\n",
    "plt.xticks(rotation=70);"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "이 방법도 정답이 아니네요. 모든 소득 분위수에서 평균 순이익이 음수입니다. 대체로 부유한 고객은 손해액이 적긴 하지만, 전체적으로 손해입니다. 지역과 같은 다른 변수는 어떨까요? 예를 들어 비용 대부분이 먼 곳의 고객에게 물건을 배송하는 배송비라면 지역은 수익성을 결정하는 `feature`가 될 것입니다.\n",
    "\n",
    "지역은 이미 범주형 변수이므로 분위수를 사용하지 않아도 됩니다. 지역별 평균 순이익을 살펴보겠습니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,6))\n",
    "np.random.seed(123)\n",
    "region_plot = sns.barplot(data=train, x=\"region\", y=\"net_value\")\n",
    "plt.title(\"Profitability by Region\");"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "빙고! $2$, $17$, $39$와 같이 수익이 나는 지역과 $0$, $9$, $29$와 같이 손해가 나는 지역을 확인할 수 있습니다. 매우 좋아 보이네요! 이제 정책을 만들 수 있습니다. **데이터에 따라** 수익이 나는 지역의 고객과만 계약합니다.\n",
    "\n",
    "지금 간단한 기계학습을 하고 있음에 주목하세요. 각 지역의 기대 순이익 $E[NetValue|Region]$을 추정하고 있습니다. 이제 추정치로 무엇인가 해봅시다.\n",
    "\n",
    "정책을 만들기 위해 지역별 예상 순수익의 $95$% 신뢰구간을 구할 것입니다. 신뢰구간이 $0$보다 크면 보다 크면 해당 지역의 고객과 계약합니다.\n",
    "\n",
    "아래 코드는 `key`가 지역(`region`)이고 `value`가 $95$% CI 하한인 `dictionary`를 만듭니다. `dictionary` 생성기는 예상 순수익이 양수인 지역만 골라냅니다. 결과는 정책에 따라 계약할 지역이 됩니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{1: 2.9729729729729737,\n",
       " 2: 20.543302704837856,\n",
       " 4: 10.051075065003388,\n",
       " 9: 32.08862469914759,\n",
       " 11: 37.434210420891255,\n",
       " 12: 37.44213667009523,\n",
       " 15: 32.09847683044394,\n",
       " 17: 39.52753893574483,\n",
       " 18: 41.86162250217046,\n",
       " 19: 15.62406327716401,\n",
       " 20: 22.06654814414531,\n",
       " 21: 24.621030401718578,\n",
       " 25: 33.97022928360584,\n",
       " 35: 11.68776141117673,\n",
       " 37: 27.83183541449011,\n",
       " 38: 49.740709395699994,\n",
       " 45: 2.286387928016998,\n",
       " 49: 17.01853709535029}"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# extract the lower bound of the 95% CI from the plot above\n",
    "regions_to_net = train.groupby('region')['net_value'].agg(['mean', 'count', 'std'])\n",
    "\n",
    "regions_to_net = regions_to_net.assign(\n",
    "    lower_bound=regions_to_net['mean'] - 1.96*regions_to_net['std']/(regions_to_net['count']**0.5)\n",
    ")\n",
    "\n",
    "regions_to_net_lower_bound = regions_to_net['lower_bound'].to_dict()\n",
    "regions_to_net = regions_to_net['mean'].to_dict()\n",
    "\n",
    "# filters regions where the net value lower bound is > 0.\n",
    "regions_to_invest = {region: net \n",
    "                     for region, net in regions_to_net_lower_bound.items()\n",
    "                     if net > 0}\n",
    "\n",
    "regions_to_invest"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`regions_to_invest`는 주목해야 할 모든 지역을 포함합니다. 이제 테스트 데이터에서 정책이 잘 작동하는지 확인해 봅시다. 정책을 평가하는 중요한 단계입니다. 우연히 훈련 데이터에 포함된 지역이 수익성이 좋아 보일 수 있기 때문입니다. 무작위성에 의해 테스트 데이터에서 훈련 데이터와 같은 패턴을 찾을 가능성은 거의 없습니다.\n",
    " \n",
    "훈련 데이터에서 수익이 높았던 지역의 고객만 포함하도록 테스트 데이터를 필터링합니다. 그리고 순이익 분포와 정책의 평균 순이익도 표시해 봅시다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "region_policy = (test[test[\"region\"]\n",
    "                      # filter regions in regions_to_invest\n",
    "                      .isin(regions_to_invest.keys())]) \n",
    "\n",
    "sns.histplot(data=region_policy, x=\"net_value\")\n",
    "# average has to be over all customers, not just the one we've filtered with the policy\n",
    "plt.title(\"Average Net Income: %.2f\" % (region_policy[\"net_value\"].sum() / test.shape[0]));"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Machine Learning Models as Policy Inputs\n",
    " \n",
    "더 잘하고 싶다면 기계학습을 사용할 수 있습니다. 기계학습은 복잡성을 가져다주지만, 미미한 개선에 그치는 경우가 대부분입니다. 하지만 상황에 따라 돈을 벌 기회를 가져다주기도 하며 이는 오늘날 기계학습이 가치 있는 이유기도 합니다.\n",
    "\n",
    "여기서는 `Gradient Boosting model`을 사용합니다. 꽤 복잡한 알고리즘이지만 간단히 사용할 수 있습니다. 우리의 목적은 $E[Y|X]$를 추정하는 것이므로 알고리즘 작동 방법을 세세하게 알 필요는 없습니다. 기계학습 모델에서 복잡성을 조절할 수 있다는 사실을 기억하시나요? 복잡할수록 모델은 강력해집니다. 하지만 너무 복잡한 모델은 과적합을 유발하며 테스트 데이터에 일반화할 수 없습니다. 따라서 모델의 복잡성을 검증하기 위해 교차검증을 해야 합니다.\n",
    "\n",
    "단순한 지역 정책을 개선하기 위한 방법을 고민해봅시다. 큰 개선은 두 가지입니다. 먼저 번거롭게 `feature`를 찾을 필요가 없습니다. 좋은 고객과 나쁜 고객을 구별할 수 있는 `feature`를 찾는 과정은 꽤 번거롭습니다. 예제는 `feature`가 $3$개로 적어 직접 찾는 것이 가능했지만 $100$개 라면 어려울 것입니다. 물론 [다중 비교](https://en.wikipedia.org/wiki/Multiple_comparisons_problem)와 거짓 양성 비율에도 주의해야 합니다. 둘째로 두 개 이상의 `feature`를 고려할 수 있게 됩니다. 데이터는 지역 외에도 고객 소득 데이터를 포함합니다. 고객의 소득만으로는 많은 것을 알 수 없었지만, 지역과 함께 고려하면 어떨까요? 해당 지역의 부유한 고객에게 집중한다면 더 많은 수익을 낼 수도 있습니다. 즉, $E[NetValue|Region, Income, Age]$가 $E[NetValue|Region]$보다 `NetValue` 예측에 더 나은 추정값이 될 수 있습니다. 많은 정보를 사용하면 순수익을 더 잘 예측할 수 있습니다.\n",
    "\n",
    "두 개 이상 `feature`를 사용한 복잡한 정책을 만들기는 매우 어려우며, `feature`의 개수에 따라 조합이 기하급수적으로 커져 실용적이지 않습니다. 대신 기계학습을 사용하면 모델에 모든 `feature`를 넣어 상호작용을 학습할 수 있습니다. \n",
    "\n",
    "모델의 목표는 `region`, `income`, `age`로 `net_value`를 예측하는 것입니다. 범주형 변수인 `region`은 숫자로 바꿔야 하는데, 훈련 데이터 내 각 지역 평균 순이익으로 대체합니다. `regions_to_net`에 값이 저장된 걸 기억하시나요? `replace` 함수를 사용해 값을 인자로 넘겨주면 `region` 값을 바꿀 수 있습니다. 이 기능은 여러 번 사용되므로 함수로 만들겠습니다. 학습이 잘되도록 `feature`를 바꾸는 과정을 `feature engineering`이라 부릅니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def encode(df): \n",
    "    return df.replace({\"region\": regions_to_net})"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "다음은 [Sklearn](https://scikit-learn.org/stable/)에서 모델을 가져옵니다. `sklearn`의 모델의 사용법은 꽤 표준화되어 있습니다. 먼저 복잡도를 지정해 모델을 만듭니다. 여기서는 `n_estimators`는 $400$, `max_depth`는 $4$로 설정합니다. `max_depth`가 깊어지고 `n_estimators`가 많아질수록 모델 성능이 좋아집니다. 물론 과적합을 고려해 너무 과하게 설정하면 안 됩니다. 다시 말하지만, 설정에 대한 자세한 내용은 알 필요가 없습니다. 매우 좋은 예측 모델이라는 점만 기억하세요. 모델을 훈련하기 위해 `feature` `X`와 `target` `net_value`로 `.fit()`을 호출합니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_params = {'n_estimators': 400,\n",
    "                'max_depth': 4,\n",
    "                'min_samples_split': 10,\n",
    "                'learning_rate': 0.01,\n",
    "                'loss': 'squared_error'}\n",
    "\n",
    "features = [\"region\", \"income\", \"age\"]\n",
    "target = \"net_value\"\n",
    "\n",
    "np.random.seed(123)\n",
    "\n",
    "reg = ensemble.GradientBoostingRegressor(**model_params)\n",
    "\n",
    "# fit model on the training set\n",
    "encoded_train = train[features].pipe(encode)\n",
    "reg.fit(encoded_train, train[target]);"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "모델 훈련이 끝났습니다. 다음으로는 **테스트** 데이터로 모델의 예측 성능을 검증합니다. 기계학습을 평가하기 위해 수많은 지표가 있습니다. 여기서는 $R^2$를 사용합니다. 이것도 자세히 알 필요는 없습니다. 연속 변수(순이익)를 예측하는 모델 평가에 유용하다는 것만 알아도 충분합니다. $R^2$는 $-\\inf$에서 $1.0$의 범위를 가집니다. 예측이 평균보다 나쁜 경우 음수입니다. $R^2$는 순이익 분산이 얼마나 모델로 설명되고 있는지 숫자로 보여줍니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train R2:  0.7108790300152951\n",
      "Test R2:  0.6938513063048141\n"
     ]
    }
   ],
   "source": [
    "train_pred = (encoded_train\n",
    "              .assign(predictions=reg.predict(encoded_train[features])))\n",
    "\n",
    "print(\"Train R2: \", r2_score(y_true=train[target], y_pred=train_pred[\"predictions\"]))\n",
    "print(\"Test R2: \", r2_score(y_true=test[target], y_pred=reg.predict(test[features].pipe(encode))))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "모델은 훈련 데이터 순수익 분산의 약 71%를 설명하고, 테스트 데이터 순이익 분산의 약 69%를 설명합니다. 모델은 훈련 데이터만 보기 때문에 훈련 데이터에서 성능이 좋은 경향이 있습니다. 과적합을 확인해 보려면 `max_depth`를 $14$로 설정해보세요. 훈련 데이터의 $R^2$는 올라가지만 테스트 데이터의 $R^2$는 낮아집니다.\n",
    "\n",
    "정책을 만들기 위해 테스트 세트의 예측값을 `prediction`에 저장합니다. 예측값은 $E[NetValue|Age, Income, Region]$ 추정치입니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>customer_id</th>\n",
       "      <th>region</th>\n",
       "      <th>income</th>\n",
       "      <th>age</th>\n",
       "      <th>net_value</th>\n",
       "      <th>prediction</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5952</th>\n",
       "      <td>5952</td>\n",
       "      <td>19</td>\n",
       "      <td>1983</td>\n",
       "      <td>23</td>\n",
       "      <td>21</td>\n",
       "      <td>47.734883</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1783</th>\n",
       "      <td>1783</td>\n",
       "      <td>31</td>\n",
       "      <td>914</td>\n",
       "      <td>31</td>\n",
       "      <td>-46</td>\n",
       "      <td>-36.026935</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4811</th>\n",
       "      <td>4811</td>\n",
       "      <td>33</td>\n",
       "      <td>1349</td>\n",
       "      <td>25</td>\n",
       "      <td>-19</td>\n",
       "      <td>22.553420</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>145</th>\n",
       "      <td>145</td>\n",
       "      <td>20</td>\n",
       "      <td>1840</td>\n",
       "      <td>26</td>\n",
       "      <td>55</td>\n",
       "      <td>48.306256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7146</th>\n",
       "      <td>7146</td>\n",
       "      <td>19</td>\n",
       "      <td>3032</td>\n",
       "      <td>34</td>\n",
       "      <td>-17</td>\n",
       "      <td>7.039414</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      customer_id  region  income  age  net_value  prediction\n",
       "5952         5952      19    1983   23         21   47.734883\n",
       "1783         1783      31     914   31        -46  -36.026935\n",
       "4811         4811      33    1349   25        -19   22.553420\n",
       "145           145      20    1840   26         55   48.306256\n",
       "7146         7146      19    3032   34        -17    7.039414"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_policy = test.assign(prediction=reg.predict(test[features].pipe(encode)))\n",
    "\n",
    "model_policy.head()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "앞에서 했던 것처럼 모델 예측값으로 평균 순이익을 구해보겠습니다. 예측값이 연속형이므로 범주형 데이터로 만들어야 합니다. 한 가지 방법은 `pandas`의 `pd.qcut`을 사용하는 것입니다. 이 함수는 데이터를 분위수로 나눕니다. 분위수로 지역 개수인 $50$을 사용해봅시다. 저는 분위수를 **모델 밴드**라 부릅니다. 개별 그룹에 대해 모델 예측 범위를 알려주기 때문입니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,6))\n",
    "\n",
    "n_bands = 50\n",
    "bands = [f\"band_{b}\" for b in range(1,n_bands+1)]\n",
    "\n",
    "np.random.seed(123)\n",
    "model_plot = sns.barplot(data=model_policy\n",
    "                         .assign(model_band = pd.qcut(model_policy[\"prediction\"], q=n_bands)),\n",
    "                         x=\"model_band\", y=\"net_value\")\n",
    "plt.title(\"Profitability by Model Prediction Quantiles\")\n",
    "plt.xticks(rotation=70);"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "순이익이 음수인 밴드와 양수인 밴드가 있으며, 잘 구별되지 않는 밴드도 존재합니다. 또한, 모델 밴드 왼쪽에서 오른쪽으로 가면서 상승 추세를 보이고 있네요.\n",
    "\n",
    "기계학습 모델로 만든 정책과 지역 데이터만 사용하는 정책을 비교하기 위해 테스트 데이터에 대해 순이익 히스토그램을 그려봅시다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "tags": [
     "hide-input"
    ]
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,6))\n",
    "model_plot_df = (model_policy[model_policy[\"prediction\"]>0])\n",
    "sns.histplot(data=model_plot_df, x=\"net_value\", color=\"C2\", label=\"model_policy\")\n",
    "\n",
    "region_plot_df = (model_policy[model_policy[\"region\"].isin(regions_to_invest.keys())])\n",
    "sns.histplot(data=region_plot_df, x=\"net_value\", label=\"region_policy\")\n",
    "\n",
    "plt.title(\"Model Net Income: %.2f;    Region Policy Net Income %.2f.\" % \n",
    "          (model_plot_df[\"net_value\"].sum() / test.shape[0],\n",
    "           region_plot_df[\"net_value\"].sum() / test.shape[0]))\n",
    "plt.legend();"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "기계학습 모델은 지역만 사용하는 단순한 정책보다 더 나은 정책을 만들어 주지만 그 효과는 미미합니다. 기계학습 정책은 고객당 $16.6$헤알, 지역 정책은 고객당 $15.5$헤알 이익을 가져다줍니다. 고객이 많아질수록 기계학습 정책으로 얻을 수 있는 총수입은 증가할 것입니다.\n",
    " \n",
    "## Fine Grain Policy\n",
    " \n",
    "앞에서 가장 단순한 정책(모든 고객을 대상)을 테스트했습니다. 이 정책은 한계 순이익을 추정하는 것으로 볼 수 있습니다. 평균적으로는 손해이므로 효과가 없어 지역을 기반으로 하는 단일 `feature`를 사용하는 정책을 만들었습니다. 정책은 $\\hat{E}[NetValue|Region] > 0$인 일부 지역에서만 사업을 하는 것으로 결과가 좋았습니다. 다음으로는 모든 `feature`를 사용해 예측 모델을 학습했습니다. 그리고 학습된 모델의 예측값이 $0$이상인 고객과만 계약하는 정책을 만들었습니다.\n",
    "\n",
    "정책이 하는 일은 매우 간단합니다. 고객과 계약하거나 하지 않는 것입니다. 지금까지 다룬 정책은 이진 분류에 해당합니다. \n",
    " \n",
    "결정의 기준이 되는 값을 **thresholding**(임계값)이라 합니다. 예측 결과가 임계값보다 큰지 작은지에 따라 다른 결정을 내리게 됩니다. 사기 탐지 모델에도 적용할 수 있습니다. 사기 탐지 모델의 예측 점수가 임계값 `X`보다 크다면 거래 승인을 반려합니다.\n",
    "\n",
    "임계값은 실제로 많은 곳에서 사용되고 있으며 이진 분류에 특히 유용합니다. 이제 상황이 더 복잡한 경우를 생각해봅시다. 예를 들어 매우 수익성이 높은 고객의 관심을 끌기 위해 광고비를 더 지출하거나 특별 대우를 해준다고 해봅시다. 이때 의사 결정은 이진 분류보다는 연속적이어야 합니다. 고객에게 얼마나 큰 비용을 투자해야 할까요?\n",
    " \n",
    "투자 비용을 결정하는 문제를 풀어봅시다. 특정 고객을 대상으로 다른 회사와 경쟁하고 있으며, 투자 비용이 많으면 고객과 계약이 성사된다고 가정합니다. 수익성에 따라 투자 비용을 다르게 하는 것이 합리적입니다. 예를 들어 수익성이 높은 고객에 많이 투자해야 하며 수익성이 낮은 고객에는 투자하면 안 됩니다.\n",
    "\n",
    "예측값을 밴드로 구분해 문제를 해결할 수 있습니다. 분위수나 크기로 구분된 밴드를 20개 만들겠습니다. 첫 번째 밴드는 하위 5% 고객을 포함합니다. *예측에 따르면* 두 번째 밴드는 하위 5% ~ 10% 고객을 포함합니다. 마지막 밴드는 가장 수익성이 높은 고객을 포함할 것입니다.\n",
    " \n",
    "구간화 역시 훈련 데이터에서 추정하고 테스트 데이터에 적용해야 합니다. 따라서 훈련 데이터에서 `pd.qcut`을 적용해 구간을 나눕니다. 구간을 나누기 위해 훈련 데이터에서 미리 계산된 결과를 전달하는 `np.digitize`를 사용합니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>customer_id</th>\n",
       "      <th>region</th>\n",
       "      <th>income</th>\n",
       "      <th>age</th>\n",
       "      <th>net_value</th>\n",
       "      <th>prediction</th>\n",
       "      <th>bands</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5952</th>\n",
       "      <td>5952</td>\n",
       "      <td>19</td>\n",
       "      <td>1983</td>\n",
       "      <td>23</td>\n",
       "      <td>21</td>\n",
       "      <td>47.734883</td>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1783</th>\n",
       "      <td>1783</td>\n",
       "      <td>31</td>\n",
       "      <td>914</td>\n",
       "      <td>31</td>\n",
       "      <td>-46</td>\n",
       "      <td>-36.026935</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4811</th>\n",
       "      <td>4811</td>\n",
       "      <td>33</td>\n",
       "      <td>1349</td>\n",
       "      <td>25</td>\n",
       "      <td>-19</td>\n",
       "      <td>22.553420</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>145</th>\n",
       "      <td>145</td>\n",
       "      <td>20</td>\n",
       "      <td>1840</td>\n",
       "      <td>26</td>\n",
       "      <td>55</td>\n",
       "      <td>48.306256</td>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7146</th>\n",
       "      <td>7146</td>\n",
       "      <td>19</td>\n",
       "      <td>3032</td>\n",
       "      <td>34</td>\n",
       "      <td>-17</td>\n",
       "      <td>7.039414</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      customer_id  region  income  age  net_value  prediction  bands\n",
       "5952         5952      19    1983   23         21   47.734883     18\n",
       "1783         1783      31     914   31        -46  -36.026935      7\n",
       "4811         4811      33    1349   25        -19   22.553420     15\n",
       "145           145      20    1840   26         55   48.306256     18\n",
       "7146         7146      19    3032   34        -17    7.039414     13"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def model_binner(prediction_column, bins):\n",
    "    # find the bins according to the training set\n",
    "    bands = pd.qcut(prediction_column, q=bins, retbins=True)[1]\n",
    "    \n",
    "    def binner_function(prediction_column):\n",
    "        return np.digitize(prediction_column, bands)\n",
    "    \n",
    "    return binner_function\n",
    "    \n",
    "\n",
    "# train the binning function\n",
    "binner_fn = model_binner(train_pred[\"predictions\"], 20)\n",
    "\n",
    "# apply the binning\n",
    "model_band = model_policy.assign(bands = binner_fn(model_policy[\"prediction\"]))\n",
    "model_band.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "tags": [
     "hide-input"
    ]
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,6))\n",
    "sns.barplot(data=model_band, x=\"bands\", y=\"net_value\")\n",
    "plt.title(\"Model Bands\");"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "$20$, $19$에 해당하는 고객에 많이 투자해야 합니다. 물론 더 많은 밴드를 만들어 미세하게 조정할 수 있습니다. 구간화를 하지 않을 수도 있는데, 모델의 예측값을 그대로 사용하되 아래와 같은 규칙을 만듭니다.\n",
    "\n",
    "```\n",
    "mkt_investments_i = model_prediction_i * 0.3\n",
    "```\n",
    "\n",
    "수식에 따라 각 고객 $i$에 대해 모델이 예측한 `net_value`의 30%의 투자합니다.\n",
    "\n",
    "## Key Ideas\n",
    " \n",
    "17장에서는 짧은 시간에 많은 것을 배웠습니다. 기계학습은 입력값을 출력값으로 매핑하는 규칙을 배움으로써 좋은 예측을 할 수 있습니다. 기계학습에서 예측은 기댓값 $E[Y|X]$을 추정하는 것과 같습니다. 또한, 예측을 통해 언어번역과 자율주행 같은 문제를 어떻게 풀 수 있는지 배웠습니다.\n",
    " \n",
    "고객을 유치하는 구체적인 예시를 통해 기계학습이 어떻게 도움이 되는지 살펴보았습니다. 어떤 고객과 거래할지 결정하는 정책을 만들기도 했습니다.\n",
    "\n",
    "**비즈니스 문제를 예측 문제로 만들 수 있다면 기계학습은 적합한 도구입니다.** 저는 이 점을 강조하고 싶습니다. 기계학습에 대한 맹신으로 사람들은 중요한 것을 잊고 쓸모없는 것을 예측하는 데 시간을 낭비하곤 합니다. 사람들은 종종 비즈니스 문제를 해결하기 위해 기계학습을 도입하기보다 예측 모델을 먼저 만들고 나서야 예측 결과로 할 수 있는 비즈니스를 찾곤 합니다. 이 방법이 효과 있을 수도 있지만, 대부분 더 큰 문제를 만듭니다.\n",
    " \n",
    "## References \n",
    " \n",
    "The things I've written here are mostly stuff from my head. I've learned them through experience. This means there isn't a direct reference I can point you to. It also means that the things I wrote here have **not** passed the academic scrutiny that good science often goes through. Instead, notice how I'm talking about things that work in practice, but I don't spend too much time explaining why that is the case. It's a sort of science from the streets, if you will. However, I am putting this up for public scrutiny, so, by all means, if you find something preposterous, open an issue and I'll address it to the best of my efforts. \n",
    " \n",
    "Finally, I believe I might have been too quick for those who were hoping for a comprehensive and detailed introduction of machine learning. To be honest, I believe that where I can truly generate value is teaching about causal inference, not machine learning. For the latter, there are tons of amazing online resources, much better than I could ever dream of creating. The classical one is [Andrew Ng's course on Machine Learning](https://www.coursera.org/learn/machine-learning) and I definitely recommend you take a look into it if you are new to machine learning.\n",
    " \n",
    "## Contribute\n",
    " \n",
    "Causal Inference for the Brave and True is an open-source material on causal inference, the statistics of science. It uses only free software, based in Python. Its goal is to be accessible monetarily and intellectually.\n",
    "If you found this book valuable and you want to support it, please go to [Patreon](https://www.patreon.com/causal_inference_for_the_brave_and_true). If you are not ready to contribute financially, you can also help by fixing typos, suggesting edits or giving feedback on passages you didn't understand. Just go to the book's repository and [open an issue](https://github.com/matheusfacure/python-causality-handbook/issues). Finally, if you liked this content, please share it with others who might find it useful and give it a [star on GitHub](https://github.com/matheusfacure/python-causality-handbook/stargazers)."
   ]
  }
 ],
 "metadata": {
  "celltoolbar": "Tags",
  "kernelspec": {
   "display_name": "pytorch",
   "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": "3.8.10"
  },
  "vscode": {
   "interpreter": {
    "hash": "65b5f243489bd9358788296533fc03025fea49f65e08ef6aa7a40b96c7113e3c"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}