bayes_search_cv.py

from skopt import BayesSearchCV
from sklearn.ensemble import GradientBoostingClassifier

from sklearn.model_selection import GridSearchCV

import matplotlib.pyplot as plt
plt.rcParams["figure.dpi"] = 100
plt.rcParams["figure.figsize"] = [10,4]

def bayes_search_cv_example():
    model = GradientBoostingClassifier(random_state=0, n_estimators=50)
    
    # Param n_iter: Number of parameter settings that are sampled; 
    # tradeoff between runtime vs quality of the solution.
    # Param cv: Number of cross validation folds, default is 3.
    # print(sklearn.metrics.SCORERS.keys())
    
    opt = BayesSearchCV(
        model,
        {
            'learning_rate':list(np.arange(0.1,0.4,0.1)),
            'n_estimators': list(range(50,200,50)),
            'min_samples_split': list(range(2,10,2))
        },
        n_iter=10,
        cv=5,
        n_jobs=-1,
        pre_dispatch="2*n_jobs",
        scoring='r2',
        random_state = 0
    )
    
    opt.fit(X_train, y_train)

    x = pd.DataFrame(opt.cv_results_)['mean_test_score'].index
    y = pd.DataFrame(opt.cv_results_)['mean_test_score']
    plot(x,y)

    print("Best Validation Score:", opt.best_score_)
    print("Test Score:", opt.score(X_test, y_test))