学生成绩分类

.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "学生成绩分类",
      "provenance": [],
      "toc_visible": true,
      "authorship_tag": "ABX9TyNDT9c9JZD+l0wV/bhTLPUK",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/gist/hyoban/fdbd75a521d27e09ed8e83870b522317/.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TXQegZc8gBFr"
      },
      "source": [
        "# 利用学生数据集预测学生成绩等级\n",
        "\n",
        "## 数据\n",
        "\n",
        "student.csv，关于部分字段的描述见决策树课件P41。数据的最后一列G3是标签。\n",
        "\n",
        "## 要求\n",
        "\n",
        "1. 使用决策树完成学生成绩等级预测，可选取部分或全部特征，分析参数对结果的影响，并进行调参优化，决策树可视化\n",
        "2. 使用两种以上常用的集成学习方法完成学生成绩等级预测，分析参数对结果的影响，并进行调参优化\n",
        "3. 对上述几种方法的结果进行比较分析，形成结论\n",
        "\n",
        "## 软件环境\n",
        "\n",
        "Jupyter Notebook\n",
        "\n",
        "## 提交形式\n",
        "\n",
        "实验报告电子版，包括文字说明和代码及结果截图，提交到多模式教学网\n",
        "\n",
        "文字描述上述过程并配过程及结果的截图，包括特征选取、数据处理、调优过程、训练得到的模型、评价结果、以及结论分析等等\n",
        "\n",
        "## 时间要求\n",
        "\n",
        "11月10日前"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KhfTob_jg_YF"
      },
      "source": [
        "## 导入使用的依赖"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nOTCYRmbMS31"
      },
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "\n",
        "from sklearn.model_selection import train_test_split, GridSearchCV\n",
        "from sklearn.preprocessing import OneHotEncoder, LabelEncoder\n",
        "from sklearn.tree import DecisionTreeClassifier\n",
        "\n",
        "from IPython.display import Image\n",
        "from sklearn import tree\n",
        "import pydotplus\n",
        "\n",
        "from sklearn.metrics import accuracy_score"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Gqg6286AhH7Y"
      },
      "source": [
        "## 定义数据处理的函数\n",
        "\n",
        "将学生成绩按照范围分级"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3DRQOu0HMp47"
      },
      "source": [
        "def grade_to_level(x):\n",
        "\tx = int(x)\n",
        "\tif x < 5:\n",
        "\t\treturn 'bad'\n",
        "\telif x >= 5 and x < 10:\n",
        "\t\treturn 'medium'\n",
        "\telif x >= 10 and x < 15:\n",
        "\t\treturn 'good'\n",
        "\telse:\n",
        "\t\treturn 'excellent'\n",
        "\n",
        "def pedu_to_level(x):\n",
        "\tx = int(x)\n",
        "\tif x > 3:\n",
        "\t\treturn 'high'\n",
        "\telif x > 1.5:\n",
        "\t\treturn 'medium'\n",
        "\telse:\n",
        "\t\treturn 'low'"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "swm_HXWZhdHx"
      },
      "source": [
        "## 从 csv 文件中读取数据"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 439
        },
        "id": "jY2Er6vRacTb",
        "outputId": "6088dafc-3adf-47a7-bcde-6afd697a5f34"
      },
      "source": [
        "stu_grade = pd.read_csv('student_1.csv')\n",
        "stu_grade"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>school</th>\n",
              "      <th>sex</th>\n",
              "      <th>address</th>\n",
              "      <th>Pstatus</th>\n",
              "      <th>Pedu</th>\n",
              "      <th>reason</th>\n",
              "      <th>guardian</th>\n",
              "      <th>traveltime</th>\n",
              "      <th>studytime</th>\n",
              "      <th>schoolsup</th>\n",
              "      <th>famsup</th>\n",
              "      <th>paid</th>\n",
              "      <th>activities</th>\n",
              "      <th>nursery</th>\n",
              "      <th>higher</th>\n",
              "      <th>internet</th>\n",
              "      <th>romantic</th>\n",
              "      <th>famrel</th>\n",
              "      <th>freetime</th>\n",
              "      <th>goout</th>\n",
              "      <th>Dalc</th>\n",
              "      <th>Walc</th>\n",
              "      <th>health</th>\n",
              "      <th>absences</th>\n",
              "      <th>G1</th>\n",
              "      <th>G2</th>\n",
              "      <th>G3</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>GP</td>\n",
              "      <td>F</td>\n",
              "      <td>U</td>\n",
              "      <td>A</td>\n",
              "      <td>4.0</td>\n",
              "      <td>course</td>\n",
              "      <td>mother</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>yes</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>6</td>\n",
              "      <td>5</td>\n",
              "      <td>6</td>\n",
              "      <td>6</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>GP</td>\n",
              "      <td>F</td>\n",
              "      <td>U</td>\n",
              "      <td>T</td>\n",
              "      <td>1.0</td>\n",
              "      <td>course</td>\n",
              "      <td>father</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>no</td>\n",
              "      <td>yes</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>no</td>\n",
              "      <td>5</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>5</td>\n",
              "      <td>5</td>\n",
              "      <td>6</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>GP</td>\n",
              "      <td>F</td>\n",
              "      <td>U</td>\n",
              "      <td>T</td>\n",
              "      <td>1.0</td>\n",
              "      <td>other</td>\n",
              "      <td>mother</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>yes</td>\n",
              "      <td>no</td>\n",
              "      <td>yes</td>\n",
              "      <td>no</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>no</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>10</td>\n",
              "      <td>7</td>\n",
              "      <td>8</td>\n",
              "      <td>10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>GP</td>\n",
              "      <td>F</td>\n",
              "      <td>U</td>\n",
              "      <td>T</td>\n",
              "      <td>3.0</td>\n",
              "      <td>home</td>\n",
              "      <td>mother</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>no</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>5</td>\n",
              "      <td>2</td>\n",
              "      <td>15</td>\n",
              "      <td>14</td>\n",
              "      <td>15</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>GP</td>\n",
              "      <td>F</td>\n",
              "      <td>U</td>\n",
              "      <td>T</td>\n",
              "      <td>3.0</td>\n",
              "      <td>home</td>\n",
              "      <td>father</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>no</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>no</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>5</td>\n",
              "      <td>4</td>\n",
              "      <td>6</td>\n",
              "      <td>10</td>\n",
              "      <td>10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>390</th>\n",
              "      <td>MS</td>\n",
              "      <td>M</td>\n",
              "      <td>U</td>\n",
              "      <td>A</td>\n",
              "      <td>2.0</td>\n",
              "      <td>course</td>\n",
              "      <td>other</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>no</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>no</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>5</td>\n",
              "      <td>5</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>5</td>\n",
              "      <td>4</td>\n",
              "      <td>11</td>\n",
              "      <td>9</td>\n",
              "      <td>9</td>\n",
              "      <td>9</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>391</th>\n",
              "      <td>MS</td>\n",
              "      <td>M</td>\n",
              "      <td>U</td>\n",
              "      <td>T</td>\n",
              "      <td>2.0</td>\n",
              "      <td>course</td>\n",
              "      <td>mother</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>no</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>5</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>14</td>\n",
              "      <td>16</td>\n",
              "      <td>16</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>392</th>\n",
              "      <td>MS</td>\n",
              "      <td>M</td>\n",
              "      <td>R</td>\n",
              "      <td>T</td>\n",
              "      <td>1.0</td>\n",
              "      <td>course</td>\n",
              "      <td>other</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>yes</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>5</td>\n",
              "      <td>5</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>10</td>\n",
              "      <td>8</td>\n",
              "      <td>7</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>393</th>\n",
              "      <td>MS</td>\n",
              "      <td>M</td>\n",
              "      <td>R</td>\n",
              "      <td>T</td>\n",
              "      <td>2.5</td>\n",
              "      <td>course</td>\n",
              "      <td>mother</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>no</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>5</td>\n",
              "      <td>0</td>\n",
              "      <td>11</td>\n",
              "      <td>12</td>\n",
              "      <td>10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>394</th>\n",
              "      <td>MS</td>\n",
              "      <td>M</td>\n",
              "      <td>U</td>\n",
              "      <td>T</td>\n",
              "      <td>1.0</td>\n",
              "      <td>course</td>\n",
              "      <td>father</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>no</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>5</td>\n",
              "      <td>5</td>\n",
              "      <td>8</td>\n",
              "      <td>9</td>\n",
              "      <td>9</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>395 rows × 27 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "    school sex address Pstatus  Pedu  ... health absences  G1  G2  G3\n",
              "0       GP   F       U       A   4.0  ...      3        6   5   6   6\n",
              "1       GP   F       U       T   1.0  ...      3        4   5   5   6\n",
              "2       GP   F       U       T   1.0  ...      3       10   7   8  10\n",
              "3       GP   F       U       T   3.0  ...      5        2  15  14  15\n",
              "4       GP   F       U       T   3.0  ...      5        4   6  10  10\n",
              "..     ...  ..     ...     ...   ...  ...    ...      ...  ..  ..  ..\n",
              "390     MS   M       U       A   2.0  ...      4       11   9   9   9\n",
              "391     MS   M       U       T   2.0  ...      2        3  14  16  16\n",
              "392     MS   M       R       T   1.0  ...      3        3  10   8   7\n",
              "393     MS   M       R       T   2.5  ...      5        0  11  12  10\n",
              "394     MS   M       U       T   1.0  ...      5        5   8   9   9\n",
              "\n",
              "[395 rows x 27 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 73
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FjXV1lMohi9e"
      },
      "source": [
        "## 对非文本数据进行编码"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        },
        "id": "SakyYzSMMsor",
        "outputId": "3dd3918a-469d-469b-b802-695f5015bce6"
      },
      "source": [
        "# 确定选取的特征\n",
        "new_data = stu_grade.iloc[:, [2, 3, 4, 8, 9, 10, 11, 14, 15, 24, 25, 26]]\n",
        "\n",
        "# 对成绩划分范围\n",
        "stu_data = new_data.copy()\n",
        "stu_data['G1'] = pd.Series(map(lambda x: grade_to_level(x), stu_data['G1']))\n",
        "stu_data['G2'] = pd.Series(map(lambda x: grade_to_level(x), stu_data['G2']))\n",
        "stu_data['G3'] = pd.Series(map(lambda x: grade_to_level(x), stu_data['G3']))\n",
        "\n",
        "# 选出需要进行数值编码的特征\n",
        "str_columns = stu_data.dtypes[stu_data.dtypes == 'object'].index\n",
        "\n",
        "# 数值编码\n",
        "for col in str_columns:\n",
        "    stu_data[col] = LabelEncoder().fit_transform(stu_data[col])\n",
        "\n",
        "# one hot 编码\n",
        "stu_data = pd.get_dummies(stu_data, columns=str_columns.drop(['G1', 'G2', 'G3']))\n",
        "\n",
        "stu_data.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>Pedu</th>\n",
              "      <th>studytime</th>\n",
              "      <th>G1</th>\n",
              "      <th>G2</th>\n",
              "      <th>G3</th>\n",
              "      <th>address_0</th>\n",
              "      <th>address_1</th>\n",
              "      <th>Pstatus_0</th>\n",
              "      <th>Pstatus_1</th>\n",
              "      <th>schoolsup_0</th>\n",
              "      <th>schoolsup_1</th>\n",
              "      <th>famsup_0</th>\n",
              "      <th>famsup_1</th>\n",
              "      <th>paid_0</th>\n",
              "      <th>paid_1</th>\n",
              "      <th>higher_0</th>\n",
              "      <th>higher_1</th>\n",
              "      <th>internet_0</th>\n",
              "      <th>internet_1</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>4.0</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1.0</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1.0</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3.0</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>3.0</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Pedu  studytime  G1  G2  ...  higher_0  higher_1  internet_0  internet_1\n",
              "0   4.0          2   3   3  ...         0         1           1           0\n",
              "1   1.0          2   3   3  ...         0         1           0           1\n",
              "2   1.0          2   3   3  ...         0         1           0           1\n",
              "3   3.0          3   1   2  ...         0         1           0           1\n",
              "4   3.0          2   3   2  ...         0         1           1           0\n",
              "\n",
              "[5 rows x 19 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 74
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NrqpDIFuh2bp"
      },
      "source": [
        "## 训练模型\n",
        "\n",
        "限制决策树的最大深度来获取更好的结果\n",
        "\n",
        "## 决策树可视化"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 436
        },
        "id": "PGlg7OzxM1ZU",
        "outputId": "3e16f772-2a23-4362-8ad0-cda82fbd6caa"
      },
      "source": [
        "X_train, X_test, y_train, y_test = train_test_split(stu_data.iloc[:, :-1], stu_data['G3'], test_size=0.3, random_state=5)\n",
        "\n",
        "dt_model = DecisionTreeClassifier(random_state=5, max_depth=2)\n",
        "dt_model.fit(X_train, y_train)\n",
        "\n",
        "dot_data = tree.export_graphviz(dt_model, \n",
        "                                out_file=None, \n",
        "                                feature_names=X_train.columns.values,\n",
        "                                class_names=['0', '1', '2', '3'],\n",
        "                                filled=True,\n",
        "                                rounded=True,\n",
        "                                special_characters=True\n",
        "                                )\n",
        "graph = pydotplus.graph_from_dot_data(dot_data)\n",
        "Image(graph.create_png())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {},
          "execution_count": 75
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JEZGmQ5ziFBO"
      },
      "source": [
        "## 判断模型得分"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "c3LTmTItN3DH",
        "outputId": "1c8363a2-9c23-42aa-dc7c-7b3d1e2b9e0e"
      },
      "source": [
        "y_pred = dt_model.predict(X_test)\n",
        "accuracy_score(y_test, y_pred)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.8823529411764706"
            ]
          },
          "metadata": {},
          "execution_count": 76
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DwrDSDeUiJtj"
      },
      "source": [
        "## 选取最佳参数"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EQYGwHELOeNG",
        "outputId": "f8768380-1e85-4bf3-d571-840bfdcd5608"
      },
      "source": [
        "entropy_thresholds = np.linspace(0, 1, 50)\n",
        "gini_thresholds = np.linspace(0, 0.5, 50)\n",
        "\n",
        "param_grid = [{'criterion': ['entropy'],\n",
        "               'min_impurity_decrease': entropy_thresholds},\n",
        "              {'criterion': ['gini'],\n",
        "               'min_impurity_decrease': gini_thresholds},\n",
        "              {'max_depth': range(2, 10)},\n",
        "              {'min_samples_split': range(2, 30, 3)}]\n",
        "\n",
        "clf = GridSearchCV(DecisionTreeClassifier(), param_grid, cv=5, return_train_score=True)\n",
        "clf.fit(X_train, y_train)\n",
        "clf.best_params_, clf.best_score_"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "({'criterion': 'entropy', 'min_impurity_decrease': 0.04081632653061224},\n",
              " 0.854935064935065)"
            ]
          },
          "metadata": {},
          "execution_count": 30
        }
      ]
    }
  ]
}