[
https://issues.apache.org/jira/browse/OPENNLP-338?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13468358#comment-13468358
]
Hyosup Shim commented on OPENNLP-338:
-------------------------------------
Hi, thanks for putting in my code.
My current code is now avoiding underflow / overflow in calculating log
likelihood.
But still there's an issue about inappropriate dealing of sparsity (zero
count in feature) in data.
And I think that i don't have enough time to solve this problem right now.
I want to solve this after the current release. How do you think about this?
Thanks.
> Add L-BFGS parameter estimation training to maxent
> --------------------------------------------------
>
> Key: OPENNLP-338
> URL: https://issues.apache.org/jira/browse/OPENNLP-338
> Project: OpenNLP
> Issue Type: New Feature
> Components: Maxent
> Reporter: Joern Kottmann
> Assignee: Hyosup Shim
> Fix For: tools-1.5.3
>
> Attachments: nl-per.testa, nl-per.testb, nl-per.train,
> patch20120814-lbfgs.txt, patch20120821-lbfgs, patch-lbfgs.txt,
> precision_curve_on_iteration.png
>
>
> Add support for the L-BFGS algorithm to train a maxent classifier.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
