基于 Python 和 Keras 实现的九层神经网络深度学习模型 和 使用 GridSearchCV 实现自动搜索最优模型参数的示例代码框架3

首先引入必要的库：

---

import numpy as np

import pandas as pd

from keras.models import Sequential

from keras.layers import Dense, Dropout

from keras.wrappers.scikit_learn import KerasClassifier

from sklearn.model_selection import GridSearchCV

---

接着定义九层神经网络的模型，这里使用了一个 Dropout 层以防止过拟合：

def create_model(optimizer='adam', init='glorot_uniform'):

    model = Sequential()

    model.add(Dense(units=64, kernel_initializer=init, activation='relu', input_dim=8))

    model.add(Dropout(0.5))

    model.add(Dense(units=64, kernel_initializer=init, activation='relu'))

    model.add(Dropout(0.5))

    model.add(Dense(units=32, kernel_initializer=init, activation='relu'))

    model.add(Dropout(0.5))

    model.add(Dense(units=16, kernel_initializer=init, activation='relu'))

    model.add(Dropout(0.5))

    model.add(Dense(units=8, kernel_initializer=init, activation='relu'))

    model.add(Dropout(0.5))

    model.add(Dense(units=4, kernel_initializer=init, activation='relu'))

    model.add(Dropout(0.5))

    model.add(Dense(units=2, kernel_initializer=init, activation='relu'))

    model.add(Dropout(0.5))

    model.add(Dense(units=1, kernel_initializer=init, activation='sigmoid'))

    model.compile(loss='binary_crossentropy', optimizer=optimizer, metrics=['accuracy'])

    return model

---

定义数据：

data = pd.read_csv('data.csv')

X = data.iloc[:, 0:8].values

y = data.iloc[:, -1].values

---

使用 GridSearchCV 进行自动搜索最优参数：

model = KerasClassifier(build_fn=create_model, epochs=10, batch_size=16, verbose=0)

optimizers = ['rmsprop', 'adam']

init = ['glorot_uniform', 'normal', 'uniform']

param_grid = dict(optimizer=optimizers, init=init)

grid = GridSearchCV(estimator=model, param_grid=param_grid)

grid_result = grid.fit(X, y)

---

最后输出最优参数和准确率：

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))

---

所有侵权请告知，立删