Dropout

代码说明：

说明：  Dropout可以作为训练深度神经网络的一种trick供选择。在每个训练批次中，通过忽略一半的特征检测器（让一半的隐层节点值为0），可以明显地减少过拟合现象。这种方式可以减少特征检测器（隐层节点）间的相互作用，检测器相互作用是指某些检测器依赖其他检测器才能发挥作用。(Dropout can be used as a trick for training depth neural networks. In each training batch, by omitting half of the feature detectors (let half of the hidden layer node value be 0), the over fitting phenomenon can be significantly reduced. This method can reduce the interaction between feature detectors (hidden layer nodes). Detector interaction means that some detectors rely on other detectors to function.)

文件列表：

NN, 0 , 2015-10-04
NN\nnapplygrads.m, 628 , 2014-05-11
NN\nnbp.m, 1638 , 2014-05-11
NN\nnchecknumgrad.m, 704 , 2014-05-11
NN\nneval.asv, 815 , 2015-10-02
NN\nneval.m, 811 , 2014-05-11
NN\nnff.m, 1849 , 2014-05-11
NN\nnpredict.m, 192 , 2014-05-11
NN\nnsetup.m, 1844 , 2014-05-11
NN\nntest.m, 184 , 2014-05-11
NN\nntrain.m, 2414 , 2014-05-11
NN\nnupdatefigures.m, 1858 , 2014-05-11