flink-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Sachin Goel (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (FLINK-1723) Add cross validation for parameter selection and validation
Date Sun, 31 May 2015 04:57:17 GMT

    [ https://issues.apache.org/jira/browse/FLINK-1723?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14566352#comment-14566352

Sachin Goel commented on FLINK-1723:

One major use of cross-validation is to determine model parameters while training. It would
be prudent to provide a framework for specifying which model parameters are to be varied and
what various values of those parameters should be used to train the model. After this, the
training phase would simply proceed as usual, essentially running the fit function for all
combinations of parameters and keeping only the best for the Prediction phase.

> Add cross validation for parameter selection and validation
> -----------------------------------------------------------
>                 Key: FLINK-1723
>                 URL: https://issues.apache.org/jira/browse/FLINK-1723
>             Project: Flink
>          Issue Type: New Feature
>          Components: Machine Learning Library
>            Reporter: Till Rohrmann
>            Assignee: Mikio Braun
>              Labels: ML
> Cross validation [1] is a standard tool to select proper parameters for you model and
to validate your results. As such it is a crucial tool for every machine learning library.
> The cross validation should work with arbitrary learners and ranges of parameters you
can specify. A first cross validation strategy it should support is the k-fold cross validation.
> Resources:
> [1] [http://en.wikipedia.org/wiki/Cross-validation]

This message was sent by Atlassian JIRA

View raw message