zoukankan      html  css  js  c++  java
  • 深度学习的Xavier初始化方法

    在tensorflow中,有一个初始化函数:tf.contrib.layers.variance_scaling_initializer。Tensorflow 官网的介绍为:

    variance_scaling_initializer(
    factor=2.0,
    mode='FAN_IN',
    uniform=False,
    seed=None,
    dtype=tf.float32
    )
    1
    2
    3
    4
    5
    6
    7
    Returns an initializer that generates tensors without scaling variance.

    When initializing a deep network, it is in principle advantageous to keep the scale of the input variance constant, so it does not explode or diminish by reaching the final layer. This initializer use the following formula:

    if mode='FAN_IN': # Count only number of input connections.
    n = fan_in
    elif mode='FAN_OUT': # Count only number of output connections.
    n = fan_out
    elif mode='FAN_AVG': # Average number of inputs and output connections.
    n = (fan_in + fan_out)/2.0

    truncated_normal(shape, 0.0, stddev=sqrt(factor / n))
    1
    2
    3
    4
    5
    6
    7
    8
    这段话可以理解为,通过使用这种初始化方法,我们能够保证输入变量的变化尺度不变,从而避免变化尺度在最后一层网络中爆炸或者弥散。

    这个方法就是 Xavier 初始化方法,可以从以下这两篇论文去了解这个方法:

    ·X. Glorot and Y. Bengio. Understanding the difficulty of training deepfeedforward neural networks. In International Conference on Artificial Intelligence and Statistics, pages 249–256, 2010.
    Y. Jia, E. Shelhamer, J. Donahue, S. Karayev, J. Long, R. Girshick, S.Guadarrama, and T. Darrell. Caffe: Convolutional architecture for fast featureembedding. arXiv:1408.5093, 2014.
    或者可以通过这些文章去了解:

    CNN数值
    三种权重的初始化方法
    深度学习——Xavier初始化方法
    ---------------------
    作者:路虽远在路上
    来源:CSDN
    原文:https://blog.csdn.net/u010185894/article/details/71104387
    版权声明:本文为博主原创文章,转载请附上博文链接!

  • 相关阅读:
    SQL SERVER没有足够的内存继续执行程序 (MSCORLIB)的解决办法
    Python之数据分析
    python之爬虫
    Linux从入门到放弃(为做一个开发+运维的全能性人才而奋斗)
    从零开始学python
    VUE学习
    Python进阶
    MySQL系列
    python基础其他
    Microsoft Office 2010/2013安装组件预设
  • 原文地址:https://www.cnblogs.com/jfdwd/p/11273984.html
Copyright © 2011-2022 走看看