[jira] [Updated] (IGNITE-9461) Implement random subspace method and provide an option to combine it with bagging

["Maxim Muzafarov (Jira)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/IGNITE-9461?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Maxim Muzafarov updated IGNITE-9461:
------------------------------------
    Fix Version/s:     (was: 2.8)
                   2.9

> Implement random subspace method and provide an option to combine it with bagging
> ---------------------------------------------------------------------------------
>
>                 Key: IGNITE-9461
>                 URL: https://issues.apache.org/jira/browse/IGNITE-9461
>             Project: Ignite
>          Issue Type: Task
>          Components: ml
>    Affects Versions: 2.6
>            Reporter: Oleg Ignatenko
>            Priority: Major
>             Fix For: 2.9
>
>
> Implement random subspace method (aka attribute bagging or feature bagging) to give ML API users more options to address overfitting. Also provide an option to combine this method with bagging.
> References:
> * [Wikipedia article|https://en.wikipedia.org/wiki/Random_subspace_method] {quote}Informally, this causes individual learners to not over-focus on features that appear highly predictive/descriptive in the training set, but fail to be as predictive for points outside that set. For this reason, random subspaces are an attractive choice for problems where the number of features is much larger than the number of training points, such as learning from fMRI data or gene expression data...{quote}
> * [Combining Bagging and Random Subspaces to Create Better Ensembles|https://pdfs.semanticscholar.org/d38f/979ad85d59fc93058279010efc73a24a712c.pdf]
> * [Bagging and the Random Subspace Method for Redundant Feature Spaces|https://link.springer.com/chapter/10.1007/3-540-48219-9_1]



--
This message was sent by Atlassian Jira
(v8.3.4#803005)