You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@systemml.apache.org by "Janardhan (JIRA)" <ji...@apache.org> on 2017/11/16 08:21:00 UTC

[jira] [Commented] (SYSTEMML-2018) Fixing Weight Decay Regularization in ADAM

    [ https://issues.apache.org/jira/browse/SYSTEMML-2018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16254926#comment-16254926 ] 

Janardhan commented on SYSTEMML-2018:
-------------------------------------

cc [~dusenberrymw]

> Fixing Weight Decay Regularization in ADAM
> ------------------------------------------
>
>                 Key: SYSTEMML-2018
>                 URL: https://issues.apache.org/jira/browse/SYSTEMML-2018
>             Project: SystemML
>          Issue Type: Improvement
>          Components: Algorithms
>            Reporter: Janardhan
>
> The common implementations of adaptive gradient algorithms, such
> as Adam, limit the potential benefit of weight decay regularization, because the
> weights do not decay multiplicatively (as would be expected for standard weight
> decay) but by an additive constant factor.
> This following paper found a way to fix regularization in Adam Optimization with one addition step(+ wx) to the gradient step :
> https://arxiv.org/pdf/1711.05101.pdf



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)