为什么损失函数越大梯度下降越快？ | realLiuSir

为什么损失函数越大梯度下降越快？

Created 2020-02-24| Updated 2021-02-26|机器学习

Word count:41|Reading time: 1 min

|Post View:

根据最基础的全连接神经网络，损失函数C是和w相关的，当C越大时，对w的求导就越大，下降就越快。

Author: realLiuSir

Link: http://yoursite.com/2020/02/24/%E4%B8%BA%E4%BB%80%E4%B9%88%E6%8D%9F%E5%A4%B1%E5%87%BD%E6%95%B0%E8%B6%8A%E5%A4%A7%E6%A2%AF%E5%BA%A6%E4%B8%8B%E9%99%8D%E8%B6%8A%E5%BF%AB%EF%BC%9F/

Copyright Notice: All articles in this blog are licensed under CC BY-NC-SA 4.0 unless stating additionally.

微信
支付寶

Recommend

信息熵_交叉熵_相对熵

为什么神经网络中参数的初始化要接近0？

Local search

Powered by hexo-generator-search