[
https://issues.apache.org/jira/browse/SPARK-9888?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Feynman Liang updated SPARK-9888:
---------------------------------
Description:
LDA has received numerous updates in 1.5, including:
* OnlineLDAOptimizer:
* Asymmetric document-topic priors
* Document-topic hyperparameter optimization
* LocalLDAModel
* predict
* logPerplexity / logLikelihood
* DistributedLDAModel:
* topDocumentsPerTopic
* topTopicsPerDoc
* Save/load
It is important to note that OnlineLDAOptimizer=>LocalLDAModel and
EMLDAOptimizer=>DistributedLDAModel now support different features. The user
guide should document these differences.
was:
LDA has received numerous updates in 1.5, including:
* OnlineLDAOptimizer:
* Asymmetric document-topic priors
* Document-topic hyperparameter optimization
* LocalLDAModel
* predict
* logPerplexity / logLikelihood
* DistributedLDAModel
* topDocumentsPerTopic
* topTopicsPerDoc
It is important to note that OnlineLDAOptimizer=>LocalLDAModel and
EMLDAOptimizer=>DistributedLDAModel now support different features. The user
guide should document these differences.
> Update LDA User Guide
> ---------------------
>
> Key: SPARK-9888
> URL: https://issues.apache.org/jira/browse/SPARK-9888
> Project: Spark
> Issue Type: Documentation
> Components: MLlib
> Reporter: Feynman Liang
> Fix For: 1.5.0
>
>
> LDA has received numerous updates in 1.5, including:
> * OnlineLDAOptimizer:
> * Asymmetric document-topic priors
> * Document-topic hyperparameter optimization
> * LocalLDAModel
> * predict
> * logPerplexity / logLikelihood
> * DistributedLDAModel:
> * topDocumentsPerTopic
> * topTopicsPerDoc
> * Save/load
> It is important to note that OnlineLDAOptimizer=>LocalLDAModel and
> EMLDAOptimizer=>DistributedLDAModel now support different features. The user
> guide should document these differences.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
