（八）sklearn神经网络

1、分类

from sklearn.neural_network import MLPClassifier
X = [[0., 0.], [1., 1.]]
y = [0, 1]
#solver : {‘lbfgs’, ‘sgd’, ‘adam’}, default ‘adam’
clf = MLPClassifier(solver='lbfgs', alpha=1e-5,
                    hidden_layer_sizes=(5, 2), random_state=1)

clf.fit(X, y)

MLPClassifier(activation='relu', alpha=1e-05, batch_size='auto', beta_1=0.9,
       beta_2=0.999, early_stopping=False, epsilon=1e-08,
       hidden_layer_sizes=(5, 2), learning_rate='constant',
       learning_rate_init=0.001, max_iter=200, momentum=0.9,
       nesterovs_momentum=True, power_t=0.5, random_state=1, shuffle=True,
       solver='lbfgs', tol=0.0001, validation_fraction=0.1, verbose=False,
       warm_start=False)

clf.predict([[2., 2.], [-1., -2.]])

array([1, 0])

[coef.shape for coef in clf.coefs_]

[(2, 5), (5, 2), (2, 1)]

#打印参数
clf.coefs_

[array([[-0.14196276, -0.02104562, -0.85522848, -3.51355396, -0.60434709],
        [-0.69744683, -0.9347486 , -0.26422217, -3.35199017,  0.06640954]]),
 array([[ 0.29164405, -0.14147894],
        [ 2.39665167, -0.6152434 ],
        [-0.51650256,  0.51452834],
        [ 4.0186541 , -0.31920293],
        [ 0.32903482,  0.64394475]]),
 array([[-4.53025854],
        [-0.86285329]])]

#打印概率
clf.predict_proba([[2., 2.], [1., 2.]])  

array([[1.96718015e-04, 9.99803282e-01],
       [1.96718015e-04, 9.99803282e-01]])

2、多分类

X = [[0., 0.], [1., 1.]]
y = [[0, 1], [1, 1]]
clf = MLPClassifier(solver='lbfgs', alpha=1e-5,
                    hidden_layer_sizes=(15,), random_state=1)

clf.fit(X, y) 

MLPClassifier(activation='relu', alpha=1e-05, batch_size='auto', beta_1=0.9,
       beta_2=0.999, early_stopping=False, epsilon=1e-08,
       hidden_layer_sizes=(15,), learning_rate='constant',
       learning_rate_init=0.001, max_iter=200, momentum=0.9,
       nesterovs_momentum=True, power_t=0.5, random_state=1, shuffle=True,
       solver='lbfgs', tol=0.0001, validation_fraction=0.1, verbose=False,
       warm_start=False)

[coef.shape for coef in clf.coefs_]

[(2, 15), (15, 2)]

clf.predict([[1., 2.]])

array([[1, 1]])

clf.predict([[0., 0.]])

array([[0, 1]])

3、回归(输出层中没有激活功能)

from sklearn.neural_network import MLPRegressor
X = [[0., 0.], [1., 1.],[2., 2.],[3,3]]
y = [0, 1,2,3]
#solver : {‘lbfgs’, ‘sgd’, ‘adam’}, default ‘adam’
clf = MLPRegressor(solver='lbfgs', alpha=1e-5,
                    hidden_layer_sizes=(5, 2), random_state=1)

clf.fit(X, y)

MLPRegressor(activation='relu', alpha=1e-05, batch_size='auto', beta_1=0.9,
       beta_2=0.999, early_stopping=False, epsilon=1e-08,
       hidden_layer_sizes=(5, 2), learning_rate='constant',
       learning_rate_init=0.001, max_iter=200, momentum=0.9,
       nesterovs_momentum=True, power_t=0.5, random_state=1, shuffle=True,
       solver='lbfgs', tol=0.0001, validation_fraction=0.1, verbose=False,
       warm_start=False)

clf.predict([[1., 1.5], [-1., 0.]])

array([ 1.36478416, -0.03321067])

4、正则化——alpha

5、实用的提示

多层感知器对特征缩放很敏感，因此需要处理数据。例如，将输入向量X上的每个属性缩放
为[0,1]或[-1，+ 1]，或将其标准化为均值0和方差1.注意，必须将相同的缩放应用于测试集。