keras与sklearn的结合使用

keras与sklearn的结合使用

新建

模板

Fly

Time: 2017-4-14

• 引言
• 代码

---

引言

众所周知，keras目前没有提供交叉验证的功能，我们要向使用交叉验证，就需要与sklearn结合。keras也提供了这样的包装接口。keras.wrappers.scikit_learn
通过这个包里面的KerasClassifier或者KerasRegressor就可以结合。闲话少叙，上代码。

代码

#!/usr/bin/python
# encoding: utf-8

"""
@version: 1.0
@author: Fly Lu
@license: Apache Licence 
@contact: luyfuyu@gmail.com
@site: https://www.flyuuu.cn
@software: PyCharm
@file: sklearn_keras.py
@time: 2017-04-09 9:23
@description: 描述sklearn使用keras
"""
from keras.wrappers.scikit_learn import KerasClassifier
from keras.models import Sequential
from keras.layers import Dense

from sklearn.cross_validation import StratifiedKFold, cross_val_score

import numpy as np


def create_model():

    model = Sequential()
    model.add(Dense(12, input_dim=8, kernel_initializer=\'uniform\', activation=\'relu\'))
    model.add(Dense(8, kernel_initializer=\'uniform\', activation=\'relu\'))
    model.add(Dense(1, kernel_initializer=\'uniform\', activation=\'sigmoid\'))
    model.compile(loss=\'binary_crossentropy\', optimizer=\'adam\', metrics=[\'accuracy\'])

    return model


# 为了让每次的结果都相同
seed = 7
np.random.seed(seed)

# 加载数据
dataset = np.loadtxt(\'./data/pima-indians-diabetes.csv\', delimiter=\',\')
X = dataset[:, 0:8]
Y = dataset[:, 8]


model = KerasClassifier(build_fn=create_model, epochs=150, batch_size=10)

kfold = StratifiedKFold(Y, n_folds=10, shuffle=True, random_state=seed)

results = cross_val_score(model, X, Y, cv=kfold)

print(np.average(results))