特征工程：交互特征与多项式特征理解

￡神魔★判官ぃ 2022-12-31 13:21 172阅读 0赞

### 特征工程：交互特征与多项式特征理解 ###

*  一、理解
 *  二、测试代码比较

# 一、理解 #

交互特征与多项式特征与数据预处理中的MinMaxScaler是相似的，都是对数据进行缩放处理  
缩放处理、交互特征与多项式特征都是对原始数据进行缩放，缩放意义在于使得权重与偏置更具有敏感性，更易对数据预测

# 二、测试代码比较 #

from sklearn.datasets import load_boston
    from sklearn.model_selection import train_test_split
    from sklearn.preprocessing import MinMaxScaler
    from sklearn.preprocessing import PolynomialFeatures
    from sklearn.linear_model import Ridge
    from sklearn.ensemble import RandomForestRegressor
    
    """波士顿房价数据集”“”
    """加载数据，利用MinMaxScaler将其放缩到0-1之间。"""
    boston = load_boston()
    X_train, X_test, y_train, y_test = train_test_split( boston.data, boston.target, random_state=0)
    scaler = MinMaxScaler()
    X_train_scaled = scaler.fit_transform(X_train)
    X_test_scaled = scaler.transform(X_test)
    
    """提取多项式特征和交互特征"""
    poly = PolynomialFeatures(degree=2).fit(X_train_scaled)
    X_train_poly = poly.transform(X_train_scaled)
    X_test_poly = poly.transform(X_test_scaled)
    print("X_train.shape: {}".format(X_train.shape))
    print("X_train_poly.shape: {}".format(X_train_poly.shape))
    
    """Ridge在有交互特征地数据集和没有交互特征地数据上地性能进行对比"""
    ridge = Ridge().fit(X_train, y_train)
    print("Origin: {:.3f}".format(ridge.score(X_test, y_test)))
    ridge = Ridge().fit(X_train_scaled, y_train)
    print("Score without interactions: {:.3f}".format(ridge.score(X_test_scaled, y_test)))
    ridge = Ridge().fit(X_train_poly, y_train)
    print("Score with interactions: {:.3f}".format(ridge.score(X_test_poly, y_test)))
    
    
    """随机森林"""
    rf = RandomForestRegressor(n_estimators=400).fit(X_train, y_train)
    print("Origin: {:.3f}".format(rf.score(X_test, y_test)))
    rf = RandomForestRegressor(n_estimators=400).fit(X_train_scaled, y_train)
    print("Score without interactions: {:.3f}".format(rf.score(X_test_scaled, y_test)))
    rf = RandomForestRegressor(n_estimators=400).fit(X_train_poly, y_train)
    print("Score with interactions: {:.3f}".format(rf.score(X_test_poly, y_test)))
    #对回归模型更能增加精确度，而对分类模型则有所相反

输出

X_train.shape: (379, 13)
    X_train_poly.shape: (379, 105)
    Origin: 0.627
    Score without interactions: 0.621
    Score with interactions: 0.753
    Origin: 0.791
    Score without interactions: 0.802
    Score with interactions: 0.768