我正在使用sklearn.cross_validation.cross_val_score函数对多层感知器进行交叉验证

from sklearn import svmimport numpy as npfrom sklearn.model_selection import cross_val_scoreclf = svm.SVC(gamma='auto')scores = cross_val_score(clf, X, np.ravel(y), cv=5, scoring='accuracy')

交叉验证正在运行，并返回了0.8579100145137881的结果。我如何更改学习率或隐藏层的数量以提高准确性？

回答：

from sklearn import svmimport numpy as npfrom sklearn.neural_network import MLPClassifierclf = MLPClassifier(solver='sgd', hidden_layer_sizes=(4,4), learning_rate_init=0.05, activation='logistic', max_iter=30000)from sklearn.model_selection import cross_val_scorescores = cross_val_score(clf, X, np.ravel(y), cv=5, scoring='accuracy')