TF-tf.nn.dropout介绍

Posted 2020-09-18 .每天进步一点点

官方的接口是这样的

tf.nn.dropout(x, keep_prob, noise_shape=None, seed=None, name=None)

根据给出的keep_prob参数，将输入tensor x按比例输出。

默认情况下, 每个元素保存或丢弃都是独立的。后面这段没太懂，以后懂了再补上： If noise_shape is specified, it must be broadcastable to the shape of x, and only dimensions with noise_shape[i] == shape(x)[i] will make independent decisions. For example, if shape(x) = [k, l, m, n] and noise_shape = [k, 1, 1, n], each batch and channel component will be kept independently and each row and column will be kept or not kept together.

x                 :  输入tensor
keep_prob    :  float类型，每个元素被保留下来的概率
noise_shape  : 一个1维的int32张量，代表了随机产生“保留/丢弃”标志的shape。
seed             : 整形变量，随机数种子。
name            : 名字，没啥用。