[
https://issues.apache.org/jira/browse/MAHOUT-9?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12608998#action_12608998
]
Grant Ingersoll commented on MAHOUT-9:
--------------------------------------
Yes, you are correct, never before seen terms are heavier, if you use the
default weights and default probability. The basic idea is they act as a
"fudge factor" when there isn't a lot of training data. I suppose, since we
are talking Map-Reduce on large data sets, one could argue that the defaults
should be much smaller, but it is easy enough to pass in your own values in
that case.
What would you propose?
> Implement MapReduce BayesianClassifier
> --------------------------------------
>
> Key: MAHOUT-9
> URL: https://issues.apache.org/jira/browse/MAHOUT-9
> Project: Mahout
> Issue Type: New Feature
> Components: Classification
> Reporter: Grant Ingersoll
> Assignee: Grant Ingersoll
> Priority: Minor
> Fix For: 0.1
>
> Attachments: MAHOUT-9.patch, MAHOUT-9.patch, MAHOUT-9.patch,
> MAHOUT-9.patch, MAHOUT-9.patch
>
>
> Implement a Bayesian classifier using M/R.
> I have a simple trainer done (not M/R) and will implement the classifier
> soon, then will upgrade it to use Hadoop.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
