votingclassifier2

votingclassifier2.py

#set parameters
params = {'voting':['hard', 'soft'],
          'weights':[(1,1,1,1,1), (2,1,1,1,1), 
                     (1,2,1,1,1), (1,1,2,1,1),
                     (1,1,1,2,1), (1,1,1,1,2), 
                     (1,1,1,2,2), (2,1,1,1,2)]}

#fit gridsearch & print best params
grid = GridSearchCV(vc, params)
grid.fit(X, y)
print('\n')
print(f'The best params is : {grid.best_params_}')

#print the final cv score
tuned_vc = VotingClassifier([('dt', DecisionTree), 
                             ('KNN', KNN), 
                             ('MLPC', MLPC), 
                             ('rf', RandomForest), 
                             ('xgb', XGB)], 
                            **grid.best_params_, n_jobs = -1)
tuned_cvm = cross_val_score(tuned_vc, X, y)
tuned_score = tuned_cvm.mean()
tuned_std = tuned_cvm.std()
print('\n')
print(f'The average tuned cross-validation score is {round(tuned_score, 4)} (+- {round(tuned_std, 4)})')