notebook/ML_loan_prediction/loan_prediction.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 银行贷款预测案例\n",
    "\n",
    "在本案例中，我们将会学习使用人工智能技术分析银行客户贷款数据，并基于这个数据训练一个高斯朴素贝叶斯分类器，预测一个新客户是否会在银行贷款。我们会展示构建银行贷款预测分类器的全流程，包括数据预处理、特征选择、模型训练和模型评估。\n",
    "\n",
    "此案例中，我们使用的数据集是银行客户信息数据集，数据集的字段描述如下：\n",
    "* ID：客户ID\n",
    "* Age：客户年龄\n",
    "* Experience：工作年限\n",
    "* Income：年度收入\n",
    "* ZIP Code：邮政编码\n",
    "* Family：家庭人数\n",
    "* CCAvg：每月信用卡消费额\n",
    "* Education：教育程度. 1: 大学未毕业; 2: 大学毕业生; 3: 研究生\n",
    "* Mortgage：房屋抵押金额\n",
    "* Personal Loan：用户是否有个人贷款?\n",
    "* Securities Account：客户在该银行中是否有证券账户?\n",
    "* CD Account：客户在该银行中是否有存款?\n",
    "* Online：用户是否会使用网络银行?\n",
    "* CreditCard：客户是否使用信用卡?\n",
    "\n",
    "本案例将掌握通过人工智能分类算法，全流程构建一个银行贷款预测分类器方案。掌握如何使用华为云ModelArts Notebook上传数据、执行Python代码。掌握如何发现并处理银行客户信息中的噪声数据。掌握如何使用Python中seaborn库绘制银行客户信息特征相关性热力图，通过相关性分析进行特征选择。掌握使用Python中matplotlib库绘制银行客户信息的基本图形，如饼图。掌握构建分类器过程中的使用的工具函数，如sklearn提供的数据集切分函数train_test_split、准确率计算函数accuracy_score。掌握使用Python编程语言实现朴素贝叶斯分类器。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 步骤"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 进入人工智能开发平台ModelArts界面\n",
    "\n",
    "这步教大家如何进入人工智能开发平台华为云ModelArts服务。\n",
    "\n",
    "第一步：点击“控制台”，如下图所示\n",
    "![title](img/enter_modelarts_step1.png)\n",
    "\n",
    "第二步：点击“所有服务”，如下图所示\n",
    "![title](img/enter_modelarts_step2.png)\n",
    "\n",
    "第三步：在“EI企业智能”大类下找到“ModelArts”，点击“ModelArts”，进入ModelArts服务主界面，如下图所示\n",
    "![title](img/enter_modelarts_step3.png)\n",
    "\n",
    "第四步：看到以下界面，说明成功进入了ModelArts服务主界面\n",
    "![title](img/enter_modelarts_step4.png)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 创建ModelArts Notebook\n",
    "\n",
    "此步教大家如何在ModelArts中创建一个Notebook开发环境。ModelArts Notebook提供网页版的Python开发环境，无需用户自己搭建Python开发环境。\n",
    "\n",
    "第一步：点击ModelArts服务主界面中的“开发环境”，如下图所示\n",
    "![title](img/create_notebook_step1.png)\n",
    "\n",
    "第二步：点击下图中的“创建”按钮\n",
    "![title](img/create_notebook_step2.png)\n",
    "\n",
    "第三步：填写创建Notebook所需的参数，并点击下一步，如下图所示\n",
    "\n",
    "\n",
    "\n",
    "| 字段 |   内容     |\n",
    "| -------- | -------------- |\n",
    "| 计费方式     | 按需计费 |\n",
    "| 名称     | Notebook实例名称，如bank_loan_prediction |\n",
    "| 工作环境 | Python3     |\n",
    "| 资源池   | 公共资源池     |\n",
    "| 类型     | CPU            |\n",
    "| 规格     | 2核 8GiB      |\n",
    "| 存储配置 | EVS，磁盘规格5GB           |\n",
    "\n",
    "\n",
    "第四步：点击下图中的“立即创建”\n",
    "![title](img/create_notebook_step4.png)\n",
    "\n",
    "第五步：点击下图中的“返回Notebook列表”\n",
    "![title](img/create_notebook_step5.png)\n",
    "\n",
    "第六步：等待Notebook创建成功，创建成功后状态会变成“运行中”，如下图所示\n",
    "![title](img/create_notebook_step6.png)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 在ModelArts Notebook中创建一个Notebook Python开发环境\n",
    "\n",
    "第一步：点击下图所示的“打开”按钮，进入刚刚创建的notebook\n",
    "![title](img/create_notebook_dev_step1.png)\n",
    "\n",
    "第二步：创建一个Notebook Python语言开发环境。先点击“New”按钮，然后创建 XGBoost-Sklearn 开发环境\n",
    "\n",
    "第三步：重命名刚刚创建的Notebook Python开发环境。点击“Untitle”，如下图所示\n",
    "![title](img/create_notebook_dev_step2.png)\n",
    "\n",
    "第四步：填写名称。我们可以填写一个跟本实验相关的名称，然后点击“Rename”按钮，如下图所示\n",
    "![title](img/create_notebook_dev_step3.png)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 如何在Notebook Python开发环境中写代码并执行代码\n",
    "\n",
    "输入代码，然后点击run按钮执行代码，就可以看到输出hello notebook!，如下图所示\n",
    "![title](img/type_code_step1.png)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Notebook Python开发环境终于准备好了，现在可以在Notebook Python开发环境写代码啦**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 案例配置信息填写\n",
    "\n",
    "案例中需要将运行结果上传至OBS中，我们需要设置以下相关参数（使用自己真实的桶名和唯一ID替换掉*号）：\n",
    "\n",
    "* BUCKET_NAME ： 自己的OBS桶名\n",
    "* UNIQUE_ID : 唯一ID，填写自己的学号或者IAM子账号名称\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "BUCKET_NAME = '*' \n",
    "UNIQUE_ID = '*' \n",
    "OBS_BASE_PATH = BUCKET_NAME + '/' + UNIQUE_ID"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 初始化ModelArts SDK"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from modelarts.session import Session\n",
    "session = Session()\n",
    "\n",
    "if session.region_name == \"cn-north-1\":\n",
    "    bucket_path = \"ai-course-common-20/bank_loan_prediction/bank_loan_prediction.tar.gz\"\n",
    "elif session.region_name == \"cn-north-4\":\n",
    "    bucket_path = \"ai-course-common-20-bj4/bank_loan_prediction/bank_loan_prediction.tar.gz\"\n",
    "else:\n",
    "    print(\"请更换地区到北京一或北京四\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 准备源代码和数据\n",
    "\n",
    "这一步准备案例所需的源代码和数据，相关资源已经保存在OBS（华为云对象存储服务）中，我们通过ModelArts SDK将资源下载到本地，并解压到当前目录下。解压后，当前目录包含src和data两个目录，分别存有源代码和数据集。\n",
    "\n",
    "会显示下载成功的日志。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Successfully download file ai-course-common-20/bank_loan_prediction/bank_loan_prediction.tar.gz from OBS to local ./bank_loan_prediction.tar.gz\n"
     ]
    }
   ],
   "source": [
    "session.download_data(bucket_path= bucket_path, path=\"./bank_loan_prediction.tar.gz\")\n",
    "\n",
    "#  使用tar命令解压资源包\n",
    "!tar xf ./bank_loan_prediction.tar.gz"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 导入基本工具库\n",
    "执行下面方框中的这段代码，可以导入本次实验中使用的Python开发基本工具库。\n",
    "\n",
    "matplotlib.pyplot是绘制基本图形的工具，seaborn绘制热力图的工具，pandas是文件读取和数据处理工具，numpy是数据分处理工具。\n",
    "\n",
    "此段代码只是引入Python包，无回显（代码执行输出）。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import numpy as np \n",
    "import pandas as pd\n",
    "# 在Notebook中显示matplotlib.pyplot的绘图结果\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 数据读取并抽取特征和预测目标\n",
    "\n",
    "使用pandas.read_excel(filepath, sheetname)方法读取Notebook中的数据文件。\n",
    "* filepath：数据文件路径。\n",
    "* sheetname: 指定数据集在Excel中所处的sheet。\n",
    "\n",
    "此段代码无回显。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "original = pd.read_excel('data/Bank_Personal_Loan_Modelling.xlsx', \"Data\")\n",
    "\n",
    "# 抽取特征数据\n",
    "feature = original.drop(\"Personal Loan\",axis=1)\n",
    "# 将数据集中的Personal Loan字段设为预测目标\n",
    "target=original[\"Personal Loan\"]\n",
    "\n",
    "# 组合成完整的数据集\n",
    "loans = feature.join(target)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 展示样本数据\n",
    "执行这段代码可以看到数据集的5个样本数据"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ID</th>\n",
       "      <th>Age</th>\n",
       "      <th>Experience</th>\n",
       "      <th>Income</th>\n",
       "      <th>ZIP Code</th>\n",
       "      <th>Family</th>\n",
       "      <th>CCAvg</th>\n",
       "      <th>Education</th>\n",
       "      <th>Mortgage</th>\n",
       "      <th>Securities Account</th>\n",
       "      <th>CD Account</th>\n",
       "      <th>Online</th>\n",
       "      <th>CreditCard</th>\n",
       "      <th>Personal Loan</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>25</td>\n",
       "      <td>1</td>\n",
       "      <td>49</td>\n",
       "      <td>91107</td>\n",
       "      <td>4</td>\n",
       "      <td>1.6</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</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>2</td>\n",
       "      <td>45</td>\n",
       "      <td>19</td>\n",
       "      <td>34</td>\n",
       "      <td>90089</td>\n",
       "      <td>3</td>\n",
       "      <td>1.5</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</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>3</td>\n",
       "      <td>39</td>\n",
       "      <td>15</td>\n",
       "      <td>11</td>\n",
       "      <td>94720</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</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>4</td>\n",
       "      <td>35</td>\n",
       "      <td>9</td>\n",
       "      <td>100</td>\n",
       "      <td>94112</td>\n",
       "      <td>1</td>\n",
       "      <td>2.7</td>\n",
       "      <td>2</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>5</td>\n",
       "      <td>35</td>\n",
       "      <td>8</td>\n",
       "      <td>45</td>\n",
       "      <td>91330</td>\n",
       "      <td>4</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   ID  Age  Experience  Income  ZIP Code  Family  CCAvg  Education  Mortgage  \\\n",
       "0   1   25           1      49     91107       4    1.6          1         0   \n",
       "1   2   45          19      34     90089       3    1.5          1         0   \n",
       "2   3   39          15      11     94720       1    1.0          1         0   \n",
       "3   4   35           9     100     94112       1    2.7          2         0   \n",
       "4   5   35           8      45     91330       4    1.0          2         0   \n",
       "\n",
       "   Securities Account  CD Account  Online  CreditCard  Personal Loan  \n",
       "0                   1           0       0           0              0  \n",
       "1                   1           0       0           0              0  \n",
       "2                   0           0       0           0              0  \n",
       "3                   0           0       0           0              0  \n",
       "4                   0           0       0           1              0  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loans.head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 展示数据集统计信息\n",
    "调用pandas.DataFrame.describe方法，可以看到各个特征的统计信息，包括样本数、均值、标准差、最小值、1/4分位数、1/2分位数、3/4分位数和最大值。transpose是矩阵的转置运算。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>ID</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>2500.500000</td>\n",
       "      <td>1443.520003</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1250.75</td>\n",
       "      <td>2500.5</td>\n",
       "      <td>3750.25</td>\n",
       "      <td>5000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Age</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>45.338400</td>\n",
       "      <td>11.463166</td>\n",
       "      <td>23.0</td>\n",
       "      <td>35.00</td>\n",
       "      <td>45.0</td>\n",
       "      <td>55.00</td>\n",
       "      <td>67.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Experience</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>20.104600</td>\n",
       "      <td>11.467954</td>\n",
       "      <td>-3.0</td>\n",
       "      <td>10.00</td>\n",
       "      <td>20.0</td>\n",
       "      <td>30.00</td>\n",
       "      <td>43.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Income</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>73.774200</td>\n",
       "      <td>46.033729</td>\n",
       "      <td>8.0</td>\n",
       "      <td>39.00</td>\n",
       "      <td>64.0</td>\n",
       "      <td>98.00</td>\n",
       "      <td>224.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ZIP Code</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>93152.503000</td>\n",
       "      <td>2121.852197</td>\n",
       "      <td>9307.0</td>\n",
       "      <td>91911.00</td>\n",
       "      <td>93437.0</td>\n",
       "      <td>94608.00</td>\n",
       "      <td>96651.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Family</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>2.396400</td>\n",
       "      <td>1.147663</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.00</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.00</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CCAvg</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>1.937913</td>\n",
       "      <td>1.747666</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.70</td>\n",
       "      <td>1.5</td>\n",
       "      <td>2.50</td>\n",
       "      <td>10.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Education</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>1.881000</td>\n",
       "      <td>0.839869</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.00</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.00</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mortgage</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>56.498800</td>\n",
       "      <td>101.713802</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>101.00</td>\n",
       "      <td>635.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Securities Account</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>0.104400</td>\n",
       "      <td>0.305809</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CD Account</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>0.060400</td>\n",
       "      <td>0.238250</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Online</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>0.596800</td>\n",
       "      <td>0.490589</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CreditCard</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>0.294000</td>\n",
       "      <td>0.455637</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Personal Loan</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>0.096000</td>\n",
       "      <td>0.294621</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     count          mean          std     min       25%  \\\n",
       "ID                  5000.0   2500.500000  1443.520003     1.0   1250.75   \n",
       "Age                 5000.0     45.338400    11.463166    23.0     35.00   \n",
       "Experience          5000.0     20.104600    11.467954    -3.0     10.00   \n",
       "Income              5000.0     73.774200    46.033729     8.0     39.00   \n",
       "ZIP Code            5000.0  93152.503000  2121.852197  9307.0  91911.00   \n",
       "Family              5000.0      2.396400     1.147663     1.0      1.00   \n",
       "CCAvg               5000.0      1.937913     1.747666     0.0      0.70   \n",
       "Education           5000.0      1.881000     0.839869     1.0      1.00   \n",
       "Mortgage            5000.0     56.498800   101.713802     0.0      0.00   \n",
       "Securities Account  5000.0      0.104400     0.305809     0.0      0.00   \n",
       "CD Account          5000.0      0.060400     0.238250     0.0      0.00   \n",
       "Online              5000.0      0.596800     0.490589     0.0      0.00   \n",
       "CreditCard          5000.0      0.294000     0.455637     0.0      0.00   \n",
       "Personal Loan       5000.0      0.096000     0.294621     0.0      0.00   \n",
       "\n",
       "                        50%       75%      max  \n",
       "ID                   2500.5   3750.25   5000.0  \n",
       "Age                    45.0     55.00     67.0  \n",
       "Experience             20.0     30.00     43.0  \n",
       "Income                 64.0     98.00    224.0  \n",
       "ZIP Code            93437.0  94608.00  96651.0  \n",
       "Family                  2.0      3.00      4.0  \n",
       "CCAvg                   1.5      2.50     10.0  \n",
       "Education               2.0      3.00      3.0  \n",
       "Mortgage                0.0    101.00    635.0  \n",
       "Securities Account      0.0      0.00      1.0  \n",
       "CD Account              0.0      0.00      1.0  \n",
       "Online                  1.0      1.00      1.0  \n",
       "CreditCard              0.0      1.00      1.0  \n",
       "Personal Loan           0.0      0.00      1.0  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loans.describe().transpose()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 特征Experience的异常值处理\n",
    "观察上面的数据集统计信息，发现特征Experience（工作年限）有负值，这是不符合常理的。我们对异常值进行处理，将值转换为其绝对值，并再次输出数据集统计信息。可以看到经过处理后，负值已经消失。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>ID</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>2500.500000</td>\n",
       "      <td>1443.520003</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1250.75</td>\n",
       "      <td>2500.5</td>\n",
       "      <td>3750.25</td>\n",
       "      <td>5000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Age</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>45.338400</td>\n",
       "      <td>11.463166</td>\n",
       "      <td>23.0</td>\n",
       "      <td>35.00</td>\n",
       "      <td>45.0</td>\n",
       "      <td>55.00</td>\n",
       "      <td>67.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Experience</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>20.134600</td>\n",
       "      <td>11.415189</td>\n",
       "      <td>0.0</td>\n",
       "      <td>10.00</td>\n",
       "      <td>20.0</td>\n",
       "      <td>30.00</td>\n",
       "      <td>43.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Income</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>73.774200</td>\n",
       "      <td>46.033729</td>\n",
       "      <td>8.0</td>\n",
       "      <td>39.00</td>\n",
       "      <td>64.0</td>\n",
       "      <td>98.00</td>\n",
       "      <td>224.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ZIP Code</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>93152.503000</td>\n",
       "      <td>2121.852197</td>\n",
       "      <td>9307.0</td>\n",
       "      <td>91911.00</td>\n",
       "      <td>93437.0</td>\n",
       "      <td>94608.00</td>\n",
       "      <td>96651.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Family</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>2.396400</td>\n",
       "      <td>1.147663</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.00</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.00</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CCAvg</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>1.937913</td>\n",
       "      <td>1.747666</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.70</td>\n",
       "      <td>1.5</td>\n",
       "      <td>2.50</td>\n",
       "      <td>10.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Education</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>1.881000</td>\n",
       "      <td>0.839869</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.00</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.00</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mortgage</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>56.498800</td>\n",
       "      <td>101.713802</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>101.00</td>\n",
       "      <td>635.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Securities Account</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>0.104400</td>\n",
       "      <td>0.305809</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CD Account</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>0.060400</td>\n",
       "      <td>0.238250</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Online</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>0.596800</td>\n",
       "      <td>0.490589</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CreditCard</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>0.294000</td>\n",
       "      <td>0.455637</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Personal Loan</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>0.096000</td>\n",
       "      <td>0.294621</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     count          mean          std     min       25%  \\\n",
       "ID                  5000.0   2500.500000  1443.520003     1.0   1250.75   \n",
       "Age                 5000.0     45.338400    11.463166    23.0     35.00   \n",
       "Experience          5000.0     20.134600    11.415189     0.0     10.00   \n",
       "Income              5000.0     73.774200    46.033729     8.0     39.00   \n",
       "ZIP Code            5000.0  93152.503000  2121.852197  9307.0  91911.00   \n",
       "Family              5000.0      2.396400     1.147663     1.0      1.00   \n",
       "CCAvg               5000.0      1.937913     1.747666     0.0      0.70   \n",
       "Education           5000.0      1.881000     0.839869     1.0      1.00   \n",
       "Mortgage            5000.0     56.498800   101.713802     0.0      0.00   \n",
       "Securities Account  5000.0      0.104400     0.305809     0.0      0.00   \n",
       "CD Account          5000.0      0.060400     0.238250     0.0      0.00   \n",
       "Online              5000.0      0.596800     0.490589     0.0      0.00   \n",
       "CreditCard          5000.0      0.294000     0.455637     0.0      0.00   \n",
       "Personal Loan       5000.0      0.096000     0.294621     0.0      0.00   \n",
       "\n",
       "                        50%       75%      max  \n",
       "ID                   2500.5   3750.25   5000.0  \n",
       "Age                    45.0     55.00     67.0  \n",
       "Experience             20.0     30.00     43.0  \n",
       "Income                 64.0     98.00    224.0  \n",
       "ZIP Code            93437.0  94608.00  96651.0  \n",
       "Family                  2.0      3.00      4.0  \n",
       "CCAvg                   1.5      2.50     10.0  \n",
       "Education               2.0      3.00      3.0  \n",
       "Mortgage                0.0    101.00    635.0  \n",
       "Securities Account      0.0      0.00      1.0  \n",
       "CD Account              0.0      0.00      1.0  \n",
       "Online                  1.0      1.00      1.0  \n",
       "CreditCard              0.0      1.00      1.0  \n",
       "Personal Loan           0.0      0.00      1.0  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 异常值处理\n",
    "loans[\"Experience\"] = loans[\"Experience\"].apply(abs)\n",
    "\n",
    "# 再次输出统计信息\n",
    "loans.describe().transpose()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 数据类型转换\n",
    "为了保证数据在计算过程中的精度，我们将数据类型转换为float64型。\n",
    "\n",
    "然后打印特征的类型信息，可以看到各个特征都转换为了float64类型。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 5000 entries, 0 to 4999\n",
      "Data columns (total 14 columns):\n",
      "ID                    5000 non-null float64\n",
      "Age                   5000 non-null float64\n",
      "Experience            5000 non-null float64\n",
      "Income                5000 non-null float64\n",
      "ZIP Code              5000 non-null float64\n",
      "Family                5000 non-null float64\n",
      "CCAvg                 5000 non-null float64\n",
      "Education             5000 non-null float64\n",
      "Mortgage              5000 non-null float64\n",
      "Securities Account    5000 non-null float64\n",
      "CD Account            5000 non-null float64\n",
      "Online                5000 non-null float64\n",
      "CreditCard            5000 non-null float64\n",
      "Personal Loan         5000 non-null float64\n",
      "dtypes: float64(14)\n",
      "memory usage: 547.0 KB\n"
     ]
    }
   ],
   "source": [
    "loans = loans.astype(float)\n",
    "\n",
    "# 展示数据类型信息\n",
    "loans.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 数据类型转换后，重新抽取特征数据和预测目标\n",
    "ID字段是一个自增的数据索引，无实际含义，所以删除。此段代码无回显。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "feature = loans.drop([\"ID\",\"Personal Loan\"],axis=1)\n",
    "target = loans[\"Personal Loan\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 展示任意两个特征之间的相关性\n",
    "\n",
    "此段代码通过seaborn绘制热力图，展示任意两个特征之间的相关性。数值的绝对值越高说明相关性越高。\n",
    "\n",
    "观察特征相关性的目的是，如果发现两个特征相关性很高，那么在特征选择时，这两个特征至少有一个特征要被丢弃。\n",
    "\n",
    "此段代码执行后，会有一张特征相关性热力图输出，可以发现Experience（工作年限）和Age（年龄）具有很高的相关性。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f32ebe5a358>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAngAAAJXCAYAAAAAbSnZAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\na1fbmgjjQZEAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f32ebd4e2e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "corr = feature.corr() # 计算任意两个特征的相关性矩阵\n",
    "\n",
    "mask = np.zeros_like(corr)\n",
    "mask[np.triu_indices_from(mask)] = True\n",
    "\n",
    "# 绘制热力图\n",
    "plt.figure(figsize=(10, 10))\n",
    "sns.heatmap(corr, mask=mask ,annot=True, square=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 展示每个特征和预测目标之间的相关性\n",
    "\n",
    "此段代码通过seaborn绘制热力图，展示每个特征和预测目标之间的相关性。数值的绝对值越高说明相关性越高，特征与预测目标之间的相关性数值越高，表示特征越重要。\n",
    "\n",
    "如果特征的重要性不高，在特征选择时需要丢弃。\n",
    "\n",
    "此段代码后会输出一张相关性热力图，可以发现Age, ZIP Code, Online, CreditCard这几个特征和预测目标相关性比较小，重要性较低。其实，从特征的含义也很好理解为什么这些特征重要性较低。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAngAAAJXCAYAAAAAbSnZAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n4DmO4ziO46xjuIHnOI7jOI6zjvH/dsO8+8RRTU8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f32e840dda0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "loans_corr = feature.join(target).corr()\n",
    "\n",
    "mask = np.zeros((13,13))\n",
    "mask[:12,:]=1\n",
    "\n",
    "plt.figure(figsize=(10, 10))\n",
    "with sns.axes_style(\"white\"):\n",
    "    sns.heatmap(loans_corr, annot=True, square=True, mask=mask)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 统计预测目标Personal Loan 正负标签值数量\n",
    "通过pandas.DataFrame内置的方法，统计正负标签值的数量。\n",
    "\n",
    "可以看到0标签的数量是4520个，1标签的数量是480个。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Labels</th>\n",
       "      <th>Personal Loan</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>4520</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>480</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Labels  Personal Loan\n",
       "0     0.0           4520\n",
       "1     1.0            480"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "source_counts = pd.DataFrame(loans[\"Personal Loan\"].value_counts()).reset_index()\n",
    "source_counts.columns =[\"Labels\",\"Personal Loan\"]\n",
    "source_counts"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 展示正负标签值分布饼状图\n",
    "通过饼状图展示正负标签各自所占的比例。\n",
    "\n",
    "可以在输出的饼状图中看到标签1所占的比例是9.6%。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAV0AAAD2CAYAAAB1JFQuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n0P8HpOzuix8cnRAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f32ebdc0908>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig1, ax1 = plt.subplots()\n",
    "# 将将第一块凸显出来，数值的大小是分割出来的与另一块的间隙\n",
    "explode = (0, 0.15)\n",
    "# 饼图的设置\n",
    "ax1.pie(source_counts[\"Personal Loan\"], explode=explode, labels=source_counts[\"Labels\"], autopct='%1.1f%%',shadow=True, startangle=70)\n",
    "ax1.axis('equal') \n",
    "# 设置图形标题\n",
    "plt.title(\"Personal Loan Percentage\")\n",
    "# 显示图形\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 各个特征的数据分布\n",
    "使用pandas.DataFrame.hist方法绘制各个特征的数据分布直方图。这是为了让大家对数据分布有一个直观的了解。\n",
    "\n",
    "可以看到12个特征的直方图分布。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[<matplotlib.axes._subplots.AxesSubplot object at 0x7f32ebcbd550>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7f32ebc72390>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7f32ebd4da58>],\n",
       "       [<matplotlib.axes._subplots.AxesSubplot object at 0x7f32e81f39b0>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7f32e8704198>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7f32e87041d0>],\n",
       "       [<matplotlib.axes._subplots.AxesSubplot object at 0x7f32e85e1400>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7f32e859ad30>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7f32e81bd9b0>],\n",
       "       [<matplotlib.axes._subplots.AxesSubplot object at 0x7f32e81cee80>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7f32e810c470>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7f32e806c198>]], dtype=object)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA7YAAANdCAYAAACgeeDZAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAACgNcEWAACA1l6eBiGPg/vH5AAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f32e84272e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# figsize可以设置图像大小\n",
    "feature.hist(figsize=(16,15))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 特征选择\n",
    "合理的特征选择可以降低模型的复杂度，提升模型的精度。\n",
    "\n",
    "基于特征相关性分析和特征与预测目标相关性分析的结果，我们丢弃掉部分不重要的特征（ZIP Code, Age, Online, CreditCard）。并且打印剩余的特征。\n",
    "\n",
    "可以看到剩余的特征有：Experience, Income, Family, CCAvg, Education, Mortgage, Securities Account, CD Account"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Index(['Experience', 'Income', 'Family', 'CCAvg', 'Education', 'Mortgage',\n",
      "       'Securities Account', 'CD Account'],\n",
      "      dtype='object')\n"
     ]
    }
   ],
   "source": [
    "feature = feature.drop([\"ZIP Code\",\"Age\",\"Online\",\"CreditCard\"],axis=1)\n",
    "\n",
    "# 打印剩余的特征\n",
    "print(feature.columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 特征缩放\n",
    "特征缩放是指将各个特征的值标准化。特征缩放的目的是消除特征值域大小对特征重要性的影响，从而让机器学习算法更加准确。\n",
    "\n",
    "我们选择\"Experience\",\"Mortgage\",\"Income\",\"CCAvg\"这几个特征来缩放，其他几个特征的值比较“标准”，无需缩放。\n",
    "\n",
    "经过StandardScaler处理后的特征具有0均值、标准差为1的特点。观察缩放之后的数据集统计信息，可以发现缩放之后的特征均值为0、标准差为1。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Experience</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>1.001421e-16</td>\n",
       "      <td>1.000100</td>\n",
       "      <td>-1.764019</td>\n",
       "      <td>-0.887906</td>\n",
       "      <td>-0.011792</td>\n",
       "      <td>0.864321</td>\n",
       "      <td>2.003268</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mortgage</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>2.810197e-16</td>\n",
       "      <td>1.000100</td>\n",
       "      <td>-0.555524</td>\n",
       "      <td>-0.555524</td>\n",
       "      <td>-0.555524</td>\n",
       "      <td>0.437558</td>\n",
       "      <td>5.688108</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Income</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>1.939449e-16</td>\n",
       "      <td>1.000100</td>\n",
       "      <td>-1.428969</td>\n",
       "      <td>-0.755483</td>\n",
       "      <td>-0.212348</td>\n",
       "      <td>0.526315</td>\n",
       "      <td>3.263712</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CCAvg</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>1.245670e-17</td>\n",
       "      <td>1.000100</td>\n",
       "      <td>-1.108969</td>\n",
       "      <td>-0.708395</td>\n",
       "      <td>-0.250595</td>\n",
       "      <td>0.321653</td>\n",
       "      <td>4.613520</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Family</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>2.396400e+00</td>\n",
       "      <td>1.147663</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>4.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Education</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>1.881000e+00</td>\n",
       "      <td>0.839869</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Securities Account</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>1.044000e-01</td>\n",
       "      <td>0.305809</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CD Account</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>6.040000e-02</td>\n",
       "      <td>0.238250</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     count          mean       std       min       25%  \\\n",
       "Experience          5000.0  1.001421e-16  1.000100 -1.764019 -0.887906   \n",
       "Mortgage            5000.0  2.810197e-16  1.000100 -0.555524 -0.555524   \n",
       "Income              5000.0  1.939449e-16  1.000100 -1.428969 -0.755483   \n",
       "CCAvg               5000.0  1.245670e-17  1.000100 -1.108969 -0.708395   \n",
       "Family              5000.0  2.396400e+00  1.147663  1.000000  1.000000   \n",
       "Education           5000.0  1.881000e+00  0.839869  1.000000  1.000000   \n",
       "Securities Account  5000.0  1.044000e-01  0.305809  0.000000  0.000000   \n",
       "CD Account          5000.0  6.040000e-02  0.238250  0.000000  0.000000   \n",
       "\n",
       "                         50%       75%       max  \n",
       "Experience         -0.011792  0.864321  2.003268  \n",
       "Mortgage           -0.555524  0.437558  5.688108  \n",
       "Income             -0.212348  0.526315  3.263712  \n",
       "CCAvg              -0.250595  0.321653  4.613520  \n",
       "Family              2.000000  3.000000  4.000000  \n",
       "Education           2.000000  3.000000  3.000000  \n",
       "Securities Account  0.000000  0.000000  1.000000  \n",
       "CD Account          0.000000  0.000000  1.000000  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.preprocessing import StandardScaler # 特征缩放类\n",
    "scaler = StandardScaler()\n",
    "\n",
    "# 参与缩放的特征\n",
    "colscal=[\"Experience\",\"Mortgage\",\"Income\",\"CCAvg\"]\n",
    "\n",
    "scaler.fit(feature[colscal])\n",
    "scaled_features = pd.DataFrame(scaler.transform(feature[colscal]),columns=colscal)\n",
    "\n",
    "feature = feature.drop(colscal, axis=1)\n",
    "feature = scaled_features.join(feature)\n",
    "\n",
    "# 展示数据集统计信息\n",
    "feature.describe().transpose()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 数据集划分\n",
    "将数据集划分为训练数据和测试数据，我们将测试数据和训练数据的比值设为3:7。返回的X_train和y_train是训练数据的特征值和标签，X_test和y_test是测试数据的特征值和标签。\n",
    "\n",
    "此段无回显。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split # 数据集切分函数\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(feature, target, test_size=0.30, random_state=101)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 查看训练数据集的样本数和特征数\n",
    "\n",
    "可以看到输出为(3500, 8)，说明训练数据集的样本数是3500个，并且特征数是8个。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3500, 8)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 导入朴素贝叶斯分类器并创建分类器对象\n",
    "\n",
    "导入的分类器是自己实现的高斯朴素贝叶斯分类器，源码见src目录下的bayes.py文件。\n",
    "\n",
    "朴素贝叶斯算法大家应该比较熟悉，但是简单的朴素贝叶斯算法只能处理离散值，无法处理连续值。而高斯朴素贝叶斯算法假设数据服从高斯分布，这样就可以处理连续值了。\n",
    "\n",
    "此段代码无回显。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 引入分类器\n",
    "from src.bayes import GaussianNB\n",
    "\n",
    "# 新建高斯朴素贝叶斯分类器对象\n",
    "clf_bayes = GaussianNB()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 训练朴素贝叶斯分类器\n",
    "\n",
    "根据训练数据训练一个高斯朴素贝叶斯分类器。源码见src目录下的bayes.py文件，大家可以查看学习。\n",
    "\n",
    "训练完成可以看到日志输出：training done。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "training done.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<src.bayes.GaussianNB at 0x7f32e74c38d0>"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_bayes.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 朴素贝叶斯模型评估\n",
    "\n",
    "根据测试集计算模型准确率，评估模型的好坏。\n",
    "\n",
    "可以看到准确率的输出，准确率是0.87。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy : 0.878666666667\n"
     ]
    }
   ],
   "source": [
    "# 引入sklearn提供的计算准确率的工具函数\n",
    "from sklearn.metrics import accuracy_score \n",
    "\n",
    "# 计算测试集的预测值\n",
    "y_predict_bayes = clf_bayes.predict(X_test)\n",
    "\n",
    "# 预测值和真实值比对，计算准确率\n",
    "print('Accuracy :',accuracy_score(y_test, y_predict_bayes))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "这是朴素贝叶斯分类器的预测结果，我们发现准确率并不是很高。接下来我们尝试训练其他的机器学习分类器，并对比效果。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 导入并创建SVM分类器\n",
    "SVM即支持向量机，是机器学习的经典算法之一。我们使用开源机器学习框架sklearn提供的SVM分类器。\n",
    "\n",
    "此段代码无回显。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.svm import SVC\n",
    "clf_svc = SVC(gamma='auto')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 训练SVM分类器\n",
    "\n",
    "代码执行完成会输出一个训练好的SVM分类器对象。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,\n",
       "  decision_function_shape=None, degree=3, gamma='auto', kernel='rbf',\n",
       "  max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
       "  tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_svc.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 评估SVM分类器\n",
    "\n",
    "可以看到SVM分类器的准确率是0.97"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy : 0.976666666667\n"
     ]
    }
   ],
   "source": [
    "# 计算测试集的预测值\n",
    "y_predict_svc = clf_svc.predict(X_test)\n",
    "\n",
    "\n",
    "# 预测值和真实值比对，计算准确率\n",
    "print('Accuracy :',accuracy_score(y_test, y_predict_svc))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 保存模型文件至OBS\n",
    "\n",
    "我们将SVM分类器训练得到的模型文件保存至OBS，方便以后再使用。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 导出模型文件\n",
    "\n",
    "训练得到模型可以导出至本地。会打印出保存的本地路径。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['results/train_model_svc.m']"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.externals import joblib\n",
    "import os\n",
    "\n",
    "if not os.path.exists('results'):\n",
    "    os.mkdir('results') # 创建本地保存路径\n",
    "\n",
    "joblib.dump(clf_svc, \"results/train_model_svc.m\") # 导出模型文件至本地"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 上传模型文件至OBS\n",
    "\n",
    "使用ModelArts SDK上传本地模型文件至OBS。可以看到上传成功的日志。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Successfully upload file ./results/train_model_svc.m to OBS xuchao3/user2/bank_loan_prediction/results\n"
     ]
    }
   ],
   "source": [
    "session.upload_data(bucket_path=OBS_BASE_PATH + '/bank_loan_prediction/results/', path='./results/train_model_svc.m')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p style=\"font-family: Arial; font-size:1.4em;color:gold;\">总结：该案例所在的OBS存储路径下，results目录下，有模型文件train_model_svc.m。</p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "通过朴素贝叶斯分类器和SVM分类器的准确率比较，发现前者比后者的准确率低了一些。虽然朴素贝叶斯分类器的准确率比其他分类器的效果要差些，但是朴素贝叶斯分类器也有优点，主要有以下几点：\n",
    "\n",
    "1.朴素贝叶斯模型发源于古典数学理论，有稳定的分类效率。\n",
    "\n",
    "2.对小规模的数据表现很好，能处理多分类任务，适合增量式训练，尤其是数据量超出内存时，我们可以一批批的去增量训练。\n",
    "\n",
    "3.对缺失数据不太敏感，算法也比较简单。"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}