[
https://issues.apache.org/jira/browse/MADLIB-1168?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16321258#comment-16321258
]
ASF GitHub Bot commented on MADLIB-1168:
----------------------------------------
GitHub user Swatisoni opened a pull request:
https://github.com/apache/madlib/pull/223
Balance datasets : re-sampling technique
JIRA:MADLIB-1168
Additional Authors:
Orhan Kislal [email protected]
Jingyi Mei [email protected]
Balanced datasets Phase 1 and Phase 2 implementation which performs
balanced sampling in following specified re-sampling techniques
1. Under-sampling the majority class(es), with- and without
replacement
2. Over-sampling the minority class
3. Combining over- and under-sampling
- Uniform sampling of all classes (default case)
4. Create ensemble balanced sets
- Re-sampling given comma-delimited string of specific class
and respective sample sizes
5. IC tests
Balanced sampling with grouping functionality will be implemented in phase 3
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/Swatisoni/madlib balanced_sets_final
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/madlib/pull/223.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #223
----
commit 3b2d1f18b9cf5ef8f78669678d82dc29cd11812b
Author: Swatisoni <soniswati.2010@...>
Date: 2018-01-10T20:07:36Z
Balance datasets : re-sampling technique
JIRA:MADLIB-1168
Additional Authors:
Orhan Kislal [email protected]
Jingyi Mei [email protected]
Balanced datasets Phase 1 and Phase 2 implementation which performs
balanced sampling in following specified re-sampling techniques
1. Under-sampling the majority class(es), with- and without
replacement
2. Over-sampling the minority class
3. Combining over- and under-sampling
- Uniform sampling of all classes (default case)
4. Create ensemble balanced sets
- Re-sampling given comma-delimited string of specific class
and respective sample sizes
5. IC tests
Balanced sampling with grouping functionality will be implemented in phase 3
----
> Balance datasets
> ----------------
>
> Key: MADLIB-1168
> URL: https://issues.apache.org/jira/browse/MADLIB-1168
> Project: Apache MADlib
> Issue Type: New Feature
> Components: Module: Sampling
> Reporter: Frank McQuillan
> Assignee: ssoni
> Fix For: v1.14
>
> Attachments: MADlib Balance Datasets Requirements.pdf,
> MADlib_Balance_Datasets_Requirements_v2.pdf
>
>
> From [1] here is the motivation behind balancing datasets:
> “Most classification algorithms will only perform optimally when the number
> of samples of each class is roughly the same. Highly skewed datasets, where
> the minority is heavily outnumbered by one or more classes, have proven to be
> a challenge while at the same time becoming more and more common.
> One way of addressing this issue is by re-sampling the dataset as to offset
> this imbalance with the hope of arriving at a more robust and fair decision
> boundary than you would otherwise.
> Re-sampling techniques can be divided in these categories:
> * Under-sampling the majority class(es).
> * Over-sampling the minority class.
> * Combining over- and under-sampling.
> * Create ensemble balanced sets.”
> There is an extensive literature on balancing datasets. The plan for MADlib
> in the initial phase is to offer basic functionality that can be extended in
> later phases based on feedback from users.
> Please see attached document for proposed scope of this story.
> References
> [1] imbalance-learn Python project
> http://contrib.scikit-learn.org/imbalanced-learn/stable/index.html
> https://github.com/scikit-learn-contrib/imbalanced-learn
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
