Feynman Liang created SPARK-9134:
------------------------------------
Summary: LDA Asymmetric topic-word prior
Key: SPARK-9134
URL: https://issues.apache.org/jira/browse/SPARK-9134
Project: Spark
Issue Type: Improvement
Components: MLlib
Reporter: Feynman Liang
SPARK-8536 generalizes LDA to asymmetric document-topic priors, which [Wallach
et al|http://dirichlet.net/pdf/wallach09rethinking.pdf] proposes offers greater
utility in terms of asymmetric priors.
However, [Stanford
NLP|http://nlp.stanford.edu/software/tmt/tmt-0.2/scaladocs/scaladocs/edu/stanford/nlp/tmt/lda/LDA.html]
also permits asymmetric priors on the topic-word prior. We should not support
manually specifying the entire matrix (which has numTopics * vocabSize
entries); rather we should follow Stanford NLP and take a single vector of
length vocabSize for a prior over words and assume that all topics share this
prior (e.g. replicate it numTopics times to get the topic-word prior matrix).
We are leaving this as todo; any users who have a need for this feature should
discuss on this JIRA.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
