ensemble learning hints

ensemble-learning-hints.py

# partial code snippets shown as hints

# voting regressor 
from sklearn.ensemble import VotingRegressor

voting_reg = VotingRegressor(
    estimators = [
        ('lin', lin_reg_pipeline),
        ('svm', svm_reg_pipeline),
        ('sgd', sgd_reg_pipeline),
    ],
)

# random forest regressor and tuning parameters

from sklearn.ensemble import RandomForestRegressor

rf_pipeline = Pipeline(
    steps=[
        ('preprocessor', preprocessor),
        ('rf', RandomForestRegressor()),
    ]
)

param_grid = [
    {
        'rf__n_estimators': [50, 100, 200, 300],
        'rf__max_depth': [2, 3, 5, 10, 20],
    }
]

# GBRT example
from sklearn.ensemble import GradientBoostingRegressor

gb_pipeline = Pipeline(
    steps=[
        ('preprocessor', preprocessor),
        ('gb', GradientBoostingRegressor()),
    ]
)