[機器學習]貝葉斯算法 對泰坦尼克號生存人群 分類預測 [簡單示例]

1.代碼

#!/usr/bin/env python
# -*- coding: utf-8 -*-
# @File  : 泰坦尼克號.py
# @Author: 趙路倉
# @Date  : 2020/4/4
# @Desc  :
# @Contact : [e-mail] 

import pandas as pd
from sklearn.model_selection import train_test_split, GridSearchCV
from sklearn.tree import DecisionTreeClassifier, export_graphviz
from sklearn.feature_extraction import DictVectorizer
from sklearn.ensemble import RandomForestClassifier



def titanic():
    # 導入數據
    titanic = pd.read_csv("titanic.txt")
    # print(titanic)

    # 篩選特征值和目標值
    x = titanic[["pclass", "age", "sex"]]
    y = titanic["survived"]
    # print(x)
    # print(y)

    # 數據處理
    # 1.缺失值處理
    x["age"].fillna(x["age"].mean(), inplace=True)
    # print(x["age"])
    # 2.轉換成字典
    x = x.to_dict(orient="records")
    # 3.數據集划分
    x_train, x_test, y_train, y_test = train_test_split(x, y, random_state=21)

    # 4.字典特征抽取
    transfer = DictVectorizer()
    x_train = transfer.fit_transform(x_train)
    x_test = transfer.transform(x_test)
    print(x_train.toarray())
    print(transfer.get_feature_names())
    # 5.決策樹預估器
    estimator = DecisionTreeClassifier(criterion="entropy")
    estimator.fit(x_train, y_train)

    # 6.模型評估
    # 方法一：直接對比真實數據和預測值
    y_predit = estimator.predict(x_test)

    print("y_predit:\n", y_predit)
    print("對比真實值和預測值:\n", y_test == y_predit)

    # 方法2：計算准確率
    score = estimator.score(x_test, y_test)
    print("准確率為：\n", score)

    # 預測
    pre=transfer.transform([{'pclass': '1st', 'age': 48.0, 'sex': 'male'}])
    prediction=estimator.predict(pre)
    print(prediction)
    # 可視化決策樹
    # 生成文件
    dot_data = export_graphviz(estimator, out_file="titanic.dot")


def titanic_forest():
    # 導入數據
    titanic = pd.read_csv("titanic.txt")
    # print(titanic)

    # 篩選特征值和目標值
    x = titanic[["pclass", "age", "sex"]]
    y = titanic["survived"]

    # 數據處理
    # 1.缺失值處理
    x["age"].fillna(x["age"].mean(), inplace=True)
    # print(x["age"])
    # 2.轉換成字典
    x = x.to_dict(orient="records")
    # 3.數據集划分
    x_train, x_test, y_train, y_test = train_test_split(x, y, random_state=21)

    # 4.字典特征抽取
    transfer = DictVectorizer()
    x_train = transfer.fit_transform(x_train)
    x_test = transfer.transform(x_test)

    estimator=RandomForestClassifier()
    # 加入網格搜索與交叉驗證
    param_dict = {
        "n_estimators": [120,200,300,500,800,1200],"max_depth":[5,8,15,25,30]
    }
    estimator = GridSearchCV(estimator, param_grid=param_dict, cv=3)
    estimator.fit(x_train, y_train)

    # 5.模型評估
    # 方法一：直接對比真實數據和預測值
    y_predit = estimator.predict(x_test)
    print("y_predit:\n", y_predit)
    print("對比真實值和預測值:\n", y_test == y_predit)

    # 方法2：計算准確率
    score = estimator.score(x_test, y_test)
    print("准確率為：\n", score)

    """
       最佳參數：best_params_
       最佳結果：best_score_
       最佳估計器：best_estimator_
       交叉驗證結果：cv_results_
    """
    print("最佳參數:\n", estimator.best_params_)
    print("最佳結果:\n", estimator.best_score_)
    print("最佳估計器:\n", estimator.best_estimator_)
    print("交叉驗證結果:\n", estimator.cv_results_)
    return None


if __name__ == "__main__":
    titanic()
    titanic_forest()

2.解釋

第一個函數 titanic() 根據游客數據

1.篩選有效數據

2.缺失值處理

3.轉換為字典

5.划分數據集

6.轉換為特征值

7.訓練模型

8.模型評估

9.預測

形成模型並評估，可以進行簡單的預測分類

第二個函數 titanic_forest() 

隨機森林找到最優方案/模型，確定最優參數等等