逻辑回归案例：电信用户流失预测 - 模型训练预测评估

import numpy as np
import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LogisticRegression
from sklearn.metrics import (
    accuracy_score,
    precision_score,
    recall_score,
    f1_score,
    classification_report
)
from sklearn.preprocessing import StandardScaler

def dm03_模型训练预测评估():
    """模型训练预测评估"""
    # 1. 加载数据集
    churn_df = pd.read_csv('data/churn.csv')
    
    # 2. 数据预处理
    # 2.1 one-hot编码
    churn_df = pd.get_dummies(churn_df, columns=['Churn', 'gender'])
    
    # 2.2 删除冗余列
    churn_df = churn_df.drop(columns=['Churn_No', 'gender_male'])
    
    # 2.3 修改列名
    churn_df = churn_df.rename(columns={'Churn_Yes': 'flag'})
    
    # 2.4 提取特征列和标签列
    # 使用的特征：月租会员、互联网服务、支付方式
    features = ['Contract_Month', 'InternetService', 'PaymentMethod_Electronic']
    X = churn_df[features]
    y = churn_df['flag']
    
    print("特征形状:", X.shape)
    print("标签分布:\n", y.value_counts())
    
    # 3. 划分训练集和测试集
    X_train, X_test, y_train, y_test = train_test_split(
        X, y, test_size=0.2, random_state=22
    )
    
    # 4. 特征工程（标准化）
    scaler = StandardScaler()
    X_train_scaled = scaler.fit_transform(X_train)
    X_test_scaled = scaler.transform(X_test)
    
    # 5. 模型训练
    model = LogisticRegression()
    model.fit(X_train_scaled, y_train)
    
    # 6. 模型预测
    y_pred = model.predict(X_test_scaled)
    
    # 7. 模型评估
    print("\n=== 模型评估 ===")
    print(f"准确率: {accuracy_score(y_test, y_pred):.4f}")
    print(f"精确率: {precision_score(y_test, y_pred):.4f}")
    print(f"召回率: {recall_score(y_test, y_pred):.4f}")
    print(f"F1值: {f1_score(y_test, y_pred):.4f}")
    
    # 8. 分类评估报告
    print("\n=== 分类评估报告 ===")
    print(classification_report(y_test, y_pred, target_names=['未流失(0)', '流失(1)']))
    
    # 9. 查看模型参数
    print("=== 模型参数 ===")
    print(f"权重W: {model.coef_}")
    print(f"偏置B: {model.intercept_}")

# 运行
if __name__ == '__main__':
    dm03_模型训练预测评估()