} model = KerasClassifier(build_fn=build_network, verbose=1) hyperparameters = create_hyperparameters() kfold_cv = KFold(n_splits=5, shuffle=True) search = RandomizedSearchCV(estimator=model, param_distributions=hyperparameters, n_iter=10, verbose=1, cv=kfold_cv) search.fit(x_train, x_train) loss, acc = search.evaluate(x_test, x_test) print(loss, acc) print('Best parameter : ', search.best_params_) print('Best estimator : ', search.best_estimator_) print('Accuracy : ', search.score(x_test, x_test)) ''' # 숫자들을 인코딩 /디코딩 # test set에서 숫자들을 가져왔다는 것을 유의 encoded_imgs = encoder.predict(x_test) decoded_imgs = decoder.predict(encoded_imgs) print(encoded_imgs) print(decoded_imgs) print(encoded_imgs.shape) # (10000, 32) print(decoded_imgs.shape) # (10000, 784)
'max_features':max_features, 'bootstrap':bootstrap} pprint(random_grid) # from sklearn.ensemble import RandomForestRegressor rf = RandomForestRegressor(random_state = 42) rf_random = RandomizedSearchCV(estimator = rf_r, param_distributions = random_grid, n_iter=100, cv = 3, verbose = 2, random_state = 42, n_jobs = -1, scoring = 'neg_mean_absolute_error') rf_random.evaluate(test_x_scaled_hhb, y_test_hhb) # error Cannot clone object '<class 'sklearn.ensemble._forest.RandomForestRegressor'>'\ # (type <class 'abc.ABCMeta'>): # it does not seem to be a scikit-learn estimator as it does not implement a 'get_params' methods. -------------------------------------------------- x_scaled_hhb test_x_scaled_hhb # 평가방법 튜닝 def evaluate(model, test_x_scaled_hhb, y_test_hhb): predictions = model.predict(test_x_scaled_hhb)