[Sklearn] 파이프라인 (pipeline)

파이프라인 (pipeline)

from sklearn.pipeline import make_pipeline

elasticnet_pipeline = make_pipeline(
    StandardScaler(),
    ElasticNet(alpha=0.1, l1_ratio=0.2)
)

elasticnet_pred = elasticnet_pipeline.fit(x_train, y_train).predict(x_test)

 

 

 

 

polynomial features

다항식의 계수간 상호작용을 통해 새로운 feature를 생성한다.

예를들면, [a, b] 2개의 feature가 존재한다고 가정하고,

degree=2로 설정한다면, polynomial features 는 [1, a, b, a^2, ab, b^2] 가 된다.

from sklearn.preprocessing import PolynomialFeatures

poly = PolynomialFeatures(degree=2, include_bias=False)

poly_features = poly.fit_transform(x_train)[0]

 

 

 

 

poly_pipeline = make_pipeline(
    PolynomialFeatures(degree=2, include_bias=False),
    StandardScaler(),
    ElasticNet(alpha=0.1, l1_ratio=0.2)
)

poly_pred = poly_pipeline.fit(x_train, y_train).predict(x_test)