[
https://issues.apache.org/jira/browse/MAHOUT-941?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Lance Norskog updated MAHOUT-941:
---------------------------------
Description:
This patch adds more statistics to the ConfusionMatrix and RequestAnalyzer.
# Add Kappa measure - a standard measure comparing a sample v.s. random
assignment.
# Add mean & standard deviation of individual labels - assist in identifying
consistent mal-assignment v.s. high and low quality labels.
Also, the SGD solver saves its model periodically to /tmp/news-groups-number.
This patch moves those captures to the model/ output directory. (These
intermediate models are interesting for tracking SGD incremental development.)
was:
This patch does 2 things:
# Add an feature to org.apache.mahout.text.SequenceFilesFromMailArchives that
removes quoted text from email bodies. This is important because it avoid
spamming the term dictionary with repeated text, especially in long email
threads.
** The feature defaults to true. Add "--quoted" to the command line to keep the
quoted lines.
# Adds some dubious overall measurements to the ConfusionMatrix.
** Kappa - a standard measurement.
*** How different is this confidence matrix from random numbers? 0.0 is the
same, 1.0 is completely different.
*** I think this is an "unweighted" kappa.
** "Success" - a homegrown formula attempting to represent the correctness of
each box. Probably bogus.
*** The standard deviation shows the distance between the success of each
producer->consumer box.
> Improve ConfusionMatrix statistics
> ----------------------------------
>
> Key: MAHOUT-941
> URL: https://issues.apache.org/jira/browse/MAHOUT-941
> Project: Mahout
> Issue Type: Improvement
> Components: Classification
> Reporter: Lance Norskog
> Assignee: Grant Ingersoll
> Priority: Minor
> Fix For: 0.7
>
> Attachments: Bayes.zip, MAHOUT-941.patch, SGD.zip
>
>
> This patch adds more statistics to the ConfusionMatrix and RequestAnalyzer.
> # Add Kappa measure - a standard measure comparing a sample v.s. random
> assignment.
> # Add mean & standard deviation of individual labels - assist in identifying
> consistent mal-assignment v.s. high and low quality labels.
> Also, the SGD solver saves its model periodically to /tmp/news-groups-number.
> This patch moves those captures to the model/ output directory. (These
> intermediate models are interesting for tracking SGD incremental development.)
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
