Why is the learning rate for the bias usually twice as large as the the LR for the weights?

1 次查看（过去 30 天）

显示更早的评论

Salma Hassan 2018-1-3

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/375451-why-is-the-learning-rate-for-the-bias-usually-twice-as-large-as-the-the-lr-for-the-weights

编辑： Image Analyst 2018-1-5

Why is the learning rate for the bias usually twice as large as the the LR for the weights?

2 个评论
显示无隐藏无

Greg Heath 2018-1-4

Please post the proof of this assertion.

Greg

Salma Hassan 2018-1-5

编辑：Image Analyst 2018-1-5

https://datascience.stackexchange.com/questions/23549/why-is-the-learning-rate-for-the-bias-usually-twice-as-large-as-the-the-lr-for-t

on the other hand if you check, for example alexnet layers you will find that net.Layers(N, 1).BiasLearnRateFactor=2 and net.Layers(N, 1).WeightLearnRateFactor = 1

请先登录，再进行评论。

请先登录，再回答此问题。

回答（0 个）

请先登录，再回答此问题。

类别

AI and Statistics Deep Learning Toolbox Image Data Workflows

在 Help Center 和 File Exchange 中查找有关 Image Data Workflows 的更多信息

产品

Deep Learning Toolbox

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by

Why is the learning rate for the bias usually twice as large as the the LR for the weights?

2 个评论
显示无隐藏无

回答（0 个）

另请参阅

类别

标签

产品

Community Treasure Hunt

Why is the learning rate for the bias usually twice as large as the the LR for the weights?

2 个评论 显示 无隐藏 无

回答（0 个）

另请参阅

类别

标签

产品

Community Treasure Hunt

2 个评论
显示无隐藏无