gridsearchCV result score가 동일하게 나옵니다.

22.03.07 16:54 작성 조회수 121

 import numpy

from sklearn.model_selection import GridSearchCV

from keras.models import Sequential

from keras.layers import Dense

from keras.wrappers.scikit_learn import KerasClassifier

#load data

data = pd.read_csv('~/all.csv')

df - data.iloc[:, 4:]

X = pd.DataFrame(df.values)

Y = pd.DataFrame(data, columns=['number'])

X_train, X_test, Y_train, Y_test = train_test_split(X.values, Y.values, test_size=0.1, random_state=42)

# Function to create model, required for KerasClassifier

def create_model(neurons=256):

 # create model

 model = Sequential()

 model.add(Dense(neurons, input_dim=8192, activation='sigmoid'))

 model.add(Dense(neurons, activation='sigmoid'))

 model.add(Dense(neurons, activation='sigmoid'))

 model.add(Dense(neurons, activation='sigmoid'))

 model.add(Dense(1, activation='linear'))

 # Compile model

 model.compile(loss='mae', optimizer='adam')

 return model

# fix random seed for reproducibility

seed = 7

numpy.random.seed(seed)

model = KerasClassifier(build_fn=create_model, verbose=0, batch_size = 8)

# define the grid search parameters

neurons = [256]

epochs = [32, 64]

param_grid = dict(neurons=neurons, epochs=epochs)

grid = GridSearchCV(estimator=model, param_grid=param_grid, n_jobs=-1, cv=3, scoring='r2')

grid_result = grid.fit(X_train, Y_train)

# summarize results

print("Best: %f using %s" % (grid_result.best_score_, grid_result.best_params_))

means = grid_result.cv_results_['mean_test_score']

stds = grid_result.cv_results_['std_test_score']

params = grid_result.cv_results_['params']

for mean, stdev, param in zip(means, stds, params):

    print("%f (%f) with: %r" % (mean, stdev, param))

코드입니다.

이렇게 실행시키면 결과값이 모두 동일한 값이 나와요.

Best: -16.710561 using {'epochs': 32, 'neurons': 256}

 -16.710561 (5.419502) with {'epochs': 32, 'neurons': 256}

 -16.710561 (5.419502) with {'epochs': 64, 'neurons': 256}

위 값 뿐만아니라 mean_test_socre 등 모든 값이 동일하게 나와 최적의 parameter를 구할수 없습니다.

무엇이 문제인지 알려주세요.

답변 0

답변을 작성해보세요.

답변을 기다리고 있는 질문이에요.
첫번째 답변을 남겨보세요!