ch04_결정트리실습_사용자행동인식.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyPiHOGC2qfvIemS1LWukDF9",
      "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/github/wlsisl/ESAA_22-2/blob/main/classification_exercise.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "N0eAMyJkJ5eS"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "###(p.201~211) 결정트리 실습 - 사용자 행동 인식 데이터 세트\n",
        "\n",
        "결정트리를 이용해 UCI머신러닝 리포지토리에서 제공하는 사용자 행동인시 ㄱ데이터 세트에 대한 얘측 분류를 수행해보자."
      ],
      "metadata": {
        "id": "tV_L3zeYJ7-0"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')\n",
        "\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline\n",
        "\n",
        "#features.txt 파일에는 피처 이름 index와 피처명이 공백으로 분리되어 있음. 이를  DataFrame으로 로드\n",
        "feature_name_df=pd.read_csv('/content/drive/MyDrive/ESAA_YB/features.txt',sep='\\s+',header=None,names=['column_index','column_name'])\n",
        "\n",
        "#피처명 index를 제거하고, 피처명만 리스트 객체로 생성한 뒤 샘플로 10개만 추출\n",
        "feature_name=feature_name_df.iloc[:,1].values.tolist()\n",
        "print('전체 피처명에서 10개만 추출:',feature_name[:10])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "OzpdFD8FJ8tz",
        "outputId": "8279678d-e552-408a-f6d4-23ed2e25f7c8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n",
            "전체 피처명에서 10개만 추출: ['tBodyAcc-mean()-X', 'tBodyAcc-mean()-Y', 'tBodyAcc-mean()-Z', 'tBodyAcc-std()-X', 'tBodyAcc-std()-Y', 'tBodyAcc-std()-Z', 'tBodyAcc-mad()-X', 'tBodyAcc-mad()-Y', 'tBodyAcc-mad()-Z', 'tBodyAcc-max()-X']\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "피처명을 보면 인체의 움직임과 관련된 속성의 평균/표준편차가 X,Y,Z축 값으로 돼 있음을 유추할 수 있다. \n",
        "\n",
        "<dataframe을 이용해 데이터 파일을 데이터 세트에 로딩하기 전 유의해야 할 부분>\n",
        "피처명을 가지고 있는 파일은 중복된 피처명이 있다. 오류가 발생하니까 변경해야함.\n"
      ],
      "metadata": {
        "id": "tn7kUUq5J8zf"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#중복된 피처가 얼마나 있는지 알아보기.\n",
        "feature_dup_df=feature_name_df.groupby('column_name').count()\n",
        "print(feature_dup_df[feature_dup_df['column_index']>1].count())\n",
        "feature_dup_df[feature_dup_df['column_index']>1].head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 274
        },
        "id": "9HbSFTGeJ830",
        "outputId": "eaf9500f-cb15-46c5-fbdf-f6d631492e8f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "column_index    42\n",
            "dtype: int64\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                              column_index\n",
              "column_name                               \n",
              "fBodyAcc-bandsEnergy()-1,16              3\n",
              "fBodyAcc-bandsEnergy()-1,24              3\n",
              "fBodyAcc-bandsEnergy()-1,8               3\n",
              "fBodyAcc-bandsEnergy()-17,24             3\n",
              "fBodyAcc-bandsEnergy()-17,32             3"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-c914eb93-508f-45e8-ba72-5c4490303db4\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>column_index</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>column_name</th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>fBodyAcc-bandsEnergy()-1,16</th>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>fBodyAcc-bandsEnergy()-1,24</th>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>fBodyAcc-bandsEnergy()-1,8</th>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>fBodyAcc-bandsEnergy()-17,24</th>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>fBodyAcc-bandsEnergy()-17,32</th>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-c914eb93-508f-45e8-ba72-5c4490303db4')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-c914eb93-508f-45e8-ba72-5c4490303db4 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-c914eb93-508f-45e8-ba72-5c4490303db4');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 2
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "결과: 42개의 피처명이 중복돼있음. \n",
        "중복된 피처명에 대해 원본 피처명에 _1 또는 _2를 추가로 부여해 새로운 피처명 가지는 DataFrame 반환하는 함수인 get_new_feature_name_df() 생성하자."
      ],
      "metadata": {
        "id": "2UvNw9ijJ88n"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "def get_new_feature_name_df(old_feature_name_df):\n",
        "  feature_dup_df=pd.DataFrame(data=old_feature_name_df.groupby('column_name').cumcount(),columns=['dup_cnt'])\n",
        "  feature_dup_df=feature_dup_df.reset_index()\n",
        "  new_feature_name_df=pd.merge(old_feature_name_df.reset_index(),feature_dup_df,how='outer')\n",
        "  new_feature_name_df['column_name']=new_feature_name_df[['column_name','dup_cnt']].apply(lambda x:x[0]+'_'+str(x[1]) if x[1]>0 else x[0],axis=1)\n",
        "  new_feature_name_df=new_feature_name_df.drop(['index'],axis=1)\n",
        "\n",
        "  return new_feature_name_df"
      ],
      "metadata": {
        "id": "qntPLHbHJ9An"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "이제 train디렉터리에 있는 학습용 피처 데이터 세트와 레이블 데이터 세트, test 디렉터리에 있는 테스트용 데이터 파일과 레이블 데이터 파일을 가갂 학습/ 테스트용 dataframe에 로드하자.\n",
        "\n"
      ],
      "metadata": {
        "id": "IZBc6k6HL3tu"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "from sklearn.ensemble import VotingClassifier\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "from sklearn.datasets import load_breast_cancer\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import accuracy_score\n",
        "\n",
        "\n",
        "def get_human_dataset( ):\n",
        "    \n",
        "    # 각 데이터 파일들은 공백으로 분리되어 있으므로 read_csv에서 공백 문자를 sep으로 할당.\n",
        "    feature_name_df = pd.read_csv('/content/drive/MyDrive/ESAA_YB/features.txt',sep='\\s+',\n",
        "                        header=None,names=['column_index','column_name'])\n",
        "    \n",
        "    # 중복된 피처명을 수정하는 get_new_feature_name_df()를 이용, 신규 피처명 DataFrame생성. \n",
        "    new_feature_name_df = get_new_feature_name_df(feature_name_df)\n",
        "    \n",
        "    # DataFrame에 피처명을 컬럼으로 부여하기 위해 리스트 객체로 다시 변환\n",
        "    feature_name = new_feature_name_df.iloc[:, 1].values.tolist()\n",
        "    \n",
        "    # 학습 피처 데이터 셋과 테스트 피처 데이터을 DataFrame으로 로딩. 컬럼명은 feature_name 적용\n",
        "    X_train = pd.read_csv('/content/drive/MyDrive/ESAA_YB/X_train.txt',sep='\\s+', names=feature_name )\n",
        "    X_test = pd.read_csv('/content/drive/MyDrive/ESAA_YB/X_test.txt',sep='\\s+', names=feature_name)\n",
        "    \n",
        "    # 학습 레이블과 테스트 레이블 데이터을 DataFrame으로 로딩하고 컬럼명은 action으로 부여\n",
        "    y_train = pd.read_csv('/content/drive/MyDrive/ESAA_YB/y_train.txt',sep='\\s+',header=None,names=['action'])\n",
        "    y_test = pd.read_csv('/content/drive/MyDrive/ESAA_YB/y_test.txt',sep='\\s+',header=None,names=['action'])\n",
        "    \n",
        "    # 로드된 학습/테스트용 DataFrame을 모두 반환 \n",
        "    return X_train, X_test, y_train, y_test\n",
        "\n",
        "X_train, X_test, y_train, y_test = get_human_dataset()\n",
        "\n",
        "print('## 학습 피처 데이터셋 info()')\n",
        "print(X_train.info())"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8xkUoMwYJ9E5",
        "outputId": "c086b1ee-c5d4-4b47-ca96-1440a1089a45"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "## 학습 피처 데이터셋 info()\n",
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 7352 entries, 0 to 7351\n",
            "Columns: 561 entries, tBodyAcc-mean()-X to angle(Z,gravityMean)\n",
            "dtypes: float64(561)\n",
            "memory usage: 31.5 MB\n",
            "None\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "학습 데이터 세트는 7352개의 레코드로 561개의 피처를 가지고 있다. 전부 float숫자형이니까 카테고리 인코딩 수행할 필요도 없어보인다."
      ],
      "metadata": {
        "id": "37oqA1hkJ9Ma"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "print(y_train['action'].value_counts())"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "HGoBrDOTJ9QZ",
        "outputId": "d766232c-5ff1-42f6-938c-8a12a99c7c3e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "6    1407\n",
            "5    1374\n",
            "4    1286\n",
            "1    1226\n",
            "2    1073\n",
            "3     986\n",
            "Name: action, dtype: int64\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "DecisionTreeClassifier : 동작 예측 분류 수행"
      ],
      "metadata": {
        "id": "m1cRycT6J9Tz"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# 하이퍼 파라미터 값 모두 추출\n",
        "\n",
        "\n",
        "from sklearn.tree import  DecisionTreeClassifier\n",
        "from sklearn.metrics import accuracy_score\n",
        "\n",
        "#예제 반복 시마다 동일한 예측 결과 도출을 위해 random_state 설정\n",
        "\n",
        "dt_clf=DecisionTreeClassifier(random_state=156)\n",
        "dt_clf.fit(X_train, y_train)\n",
        "pred=dt_clf.predict(X_test)\n",
        "accuracy=accuracy_score(y_test,pred)\n",
        "print('결정 트리 예측 정확도:{0:.4f}'.format(accuracy))\n",
        "\n",
        "#DecisionTreeClassifier의 하이퍼 파라미터 추출\n",
        "print('DecisionTreeClassifier 기본 하이퍼 파라미터:\\n',dt_clf.fit)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "C3LZADNGJ9Xq",
        "outputId": "be72d397-6836-4152-db96-86b39f4b74f2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "결정 트리 예측 정확도:0.8548\n",
            "DecisionTreeClassifier 기본 하이퍼 파라미터:\n",
            " <bound method DecisionTreeClassifier.fit of DecisionTreeClassifier(random_state=156)>\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "결과: 약 85.48의 정확도를 나타낸다.\n",
        "\n",
        "### 결정트리의 트리 깊이가 예측 정확도에 주는 영향\n",
        "\n",
        "* 결정트리의 경우 분류를 위해 리프 노드가 될 수 있는 적합한 수준이 될 때까지 지속해서 트리의 분할을 수행하면서 깊이가 깊어진다.\n",
        "* GridSearchCV를 이용해 사이킷런 결정 트리의 깊이를 조절할 수 있는 하이퍼 파라미터인 max_depth값을 변화시키면서 예측 성능 확인하기."
      ],
      "metadata": {
        "id": "pgYCWpwSJ9a7"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.model_selection import GridSearchCV\n",
        "\n",
        "params={'max_depth':[6,8,10,12,16,20,24]}\n",
        "\n",
        "grid_cv=GridSearchCV(dt_clf,param_grid=params,scoring='accuracy',cv=5,verbose=1)\n",
        "grid_cv.fit(X_train, y_train)\n",
        "print('GridSearchCV 최고 평균 정확도 수치:{0:.4f}'.format(grid_cv.best_score_))\n",
        "print('GrindSearchCV 최적 하이퍼 파라미터:',grid_cv.best_params_)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "OOIuNZoyJ9ef",
        "outputId": "33a6d1dc-feef-4895-a62d-99ca1dec5044"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Fitting 5 folds for each of 7 candidates, totalling 35 fits\n",
            "GridSearchCV 최고 평균 정확도 수치:0.8513\n",
            "GrindSearchCV 최적 하이퍼 파라미터: {'max_depth': 16}\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "결과: max_depth가 8일 때 5개의 폴드 세트의 최고 평균 정확도 결과가 85.26%로 도출.\n"
      ],
      "metadata": {
        "id": "BTCKpdbMJ9ig"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "#GridSearchCV 객체의 cv_results_ 속성을 DataFrame으로 생성\n",
        "cv_results_df=pd.DataFrame(grid_cv.cv_results_)\n",
        "\n",
        "#max_depth 파라미터 값과 그때의 테스트 세트, 학습 데이터 세트의 정확도 수치 추출\n",
        "cv_results_df[['param_max_depth','mean_test_score']]\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 270
        },
        "id": "QXZInHQ1RD0n",
        "outputId": "4b086f1b-45a1-4096-980d-880931e10a43"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "  param_max_depth  mean_test_score\n",
              "0               6         0.850791\n",
              "1               8         0.851069\n",
              "2              10         0.851209\n",
              "3              12         0.844135\n",
              "4              16         0.851344\n",
              "5              20         0.850800\n",
              "6              24         0.849440"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-2eff8acb-33ff-4f35-9747-511488ae33c8\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>param_max_depth</th>\n",
              "      <th>mean_test_score</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>6</td>\n",
              "      <td>0.850791</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>8</td>\n",
              "      <td>0.851069</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>10</td>\n",
              "      <td>0.851209</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>12</td>\n",
              "      <td>0.844135</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>16</td>\n",
              "      <td>0.851344</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>20</td>\n",
              "      <td>0.850800</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>24</td>\n",
              "      <td>0.849440</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-2eff8acb-33ff-4f35-9747-511488ae33c8')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-2eff8acb-33ff-4f35-9747-511488ae33c8 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-2eff8acb-33ff-4f35-9747-511488ae33c8');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 33
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "mean_test_score는 5개 CV세트에서 검증용 데이터 세트의 정확도 평균 수치이다.\n",
        "\n",
        "mean_test_score는 max_depth가 8일 때 0.852로 정확도가 정점이고 그 뒤로 떨어짐.\n",
        "\n",
        "-> 결정 트리는 더 완벽한 규칙을 학습 데이터 세트에 적용하기 위해 노드를 지속적으로 분할하면서 깊이가 깊어지고 더욱  복잡한 모델이 된다. \n",
        " \n",
        " 올바른 예측 결과 but 과적합으로 성능 저하\n",
        "\n"
      ],
      "metadata": {
        "id": "iaceUiJ5RD9K"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "#테스트 데이터 세트에서 결정 트리의 정확도 측정. max_depth의 변화에 따른 값 측정\n",
        "max_depths=[6,8,10,12,16,20,24]\n",
        "\n",
        "#max_depth 값을 변화시키면서 그때마다 학습과 테스트 세트에서의 예측 성능 측정\n",
        "\n",
        "for depth in max_depths:\n",
        "  dt_clf=DecisionTreeClassifier(max_depth=depth,random_state=156)\n",
        "  dt_clf.fit(X_train,y_train)\n",
        "  pred=dt_clf.predict(X_test)\n",
        "  accuracy=accuracy_score(y_test,pred)\n",
        "  print('max_depth={0} 정확도:{1:.4f}'.format(depth,accuracy))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "s7xEjVvEREC6",
        "outputId": "78c56c53-ee7c-4c21-9bdf-6f9a8f1f0fec"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "max_depth=6 정확도:0.8558\n",
            "max_depth=8 정확도:0.8707\n",
            "max_depth=10 정확도:0.8673\n",
            "max_depth=12 정확도:0.8646\n",
            "max_depth=16 정확도:0.8575\n",
            "max_depth=20 정확도:0.8548\n",
            "max_depth=24 정확도:0.8548\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "max_depth가 8일 경우 가장 높은 정확도를 나타낸다.\n",
        "\n",
        "하지만 깊이가 깊어져서 정확도는 떨어지고 있으므로 복잡한 모델보다도 트리 깊이를 낮춘 단순한 모델이 효과적인 결과를 가져올 수 있음\n",
        "\n",
        "해결: max_depth와 min_samples_split을 같이 변경하면서 정확도 성능으 튜닝하자."
      ],
      "metadata": {
        "id": "t9cPGjsVREH8"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "params={\n",
        "    'max_depth':[8,12,16,20],\n",
        "    'min_samples_split':[16,24],\n",
        "}\n",
        "\n",
        "\n",
        "grid_cv=GridSearchCV(dt_clf,param_grid=params,scoring='accuracy',cv=5,verbose=1)\n",
        "grid_cv.fit(X_train,y_train)\n",
        "print('GridSearchCV 최고 평균 정확도 수치:{0:.4f}'.format(grid_cv.best_score_))\n",
        "print('GridSearchCV 최적 하이퍼 파라미터:',grid_cv.best_params_)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "awmJsdCNREMT",
        "outputId": "e033c8c7-4f4c-4bf2-86ab-130f4e1d11db"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Fitting 5 folds for each of 8 candidates, totalling 40 fits\n",
            "GridSearchCV 최고 평균 정확도 수치:0.8549\n",
            "GridSearchCV 최적 하이퍼 파라미터: {'max_depth': 8, 'min_samples_split': 16}\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "별도로 분리된 테스트 데이터 세트에 해당 하이퍼 파라미터를 적용해보자. "
      ],
      "metadata": {
        "id": "pkveZdRnREQ_"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#테스트 데이터 세트에 예측 수행\n",
        "best_df_clf=grid_cv.best_estimator_\n",
        "pred1=best_df_clf.predict(X_test)\n",
        "accuracy=accuracy_score(y_test,pred1)\n",
        "print('결정 트리 예측 정확도:{0:.4f}'.format(accuracy))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UuPM2kuxREVB",
        "outputId": "39819388-615f-485f-8661-5d3b655b5447"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "결정 트리 예측 정확도:0.8717\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "이제 결정 트리에서 각 피처의 중요도를 feature_importances_속성으 ㄹ이용해 알아보자."
      ],
      "metadata": {
        "id": "zjxO42flSRww"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#중요도 높은 순으로 Top20 피처를 막대기로 표현\n",
        "\n",
        "import seaborn as sns\n",
        "\n",
        "frt_importances_values=best_df_clf.feature_importances_\n",
        "\n",
        "#Top  중요도로 정렬을 쉽게 하고, 시본의 막대그래프로 쉽게 표현하기 위해 Series 반환\n",
        "frt_importances=pd.Series(frt_importances_values,index=X_train.columns)\n",
        "#중요도 값 순으로 Series 를 정렬\n",
        "ftr_top20=frt_importances.sort_values(ascending=False)[:20]\n",
        "plt.figure(figsize=(8,6))\n",
        "plt.title('Feature importances Top 20')\n",
        "sns.barplot(x=ftr_top20,y=ftr_top20.index)\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Cv_ip4Q5SSBA",
        "outputId": "6aedc7b2-b214-4e53-b29a-2ce8e6cded2a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x432 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "-> 이 중 가장높은 중요도를 가진 피처들이 매우 중요하게 규칙 생성에 영향을 미치고 있음을 알 수 있다."
      ],
      "metadata": {
        "id": "bdDyWMH3SSEe"
      }
    }
  ]
}