[
https://issues.apache.org/jira/browse/HIVEMALL-22?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15714414#comment-15714414
]
ASF GitHub Bot commented on HIVEMALL-22:
----------------------------------------
GitHub user myui opened a pull request:
https://github.com/apache/incubator-hivemall/pull/4
[WIP] Kernelized Passive-Aggressive Algorithm for Classification
This PR introduce the following algorithms:
- Passive-Aggressive with Polynomial Kernel
- Polynomial Kernel with Inverted Indices (PKI)
- Kernel Expansion
- PA-I, PA-II
This PR is based on [a pending
PR](https://github.com/myui/hivemall/pull/304) by @L3Sota that is sent before
Hivemall entered Apache Incubator.
See [JIRA](https://issues.apache.org/jira/browse/HIVEMALL-22
) for tracking the status of this issue.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/myui/incubator-hivemall JIRA-22/pr-304
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/incubator-hivemall/pull/4.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 #4
----
commit 4cb84773f4c88f757d70b480a80068d597e756a7
Author: myui <yuin...@gmail.com>
Date: 2016-06-16T22:59:11Z
Changed the behavior of binarize_label
commit adb23e3c537d9c2d2ce36c27700a85ff6852964b
Author: Sotaro Sugimoto <sotaro.sugim...@gmail.com>
Date: 2016-04-12T06:23:12Z
Added Kernelized Passive-Aggressive Algorithm for Classification Problems.
commit bc228746c7f7b1abc0f6e7846d52ca1edad7fea7
Author: Sotaro Sugimoto <sotaro.sugim...@gmail.com>
Date: 2016-04-12T09:14:45Z
Bug fixes
Also newly added KernelizedPassiveAggressiveUDTFTest based on
PassiveAggressiveUDTFTest, but tests are not done for newly added functionality.
commit adac4e5bd0d89cc8c68768c9fb5bc1765bc3fac3
Author: Sotaro Sugimoto <sotaro.sugim...@gmail.com>
Date: 2016-04-14T09:51:14Z
Test for news20 written; some bugs fixed.
Bug still occurs for news20-small.binary where the loss does not match
between udtf and udtfPKI starting at sample 51, where the label switches from
all -1s to all +1s. Also, news20.binary not included due to large filesize.
commit 6072a2ae7f1d0f8a0023dbe428a38c201386e38e
Author: myui <yuin...@gmail.com>
Date: 2016-04-14T11:51:57Z
Refactored to use Bitset for support vector indices
commit ecf4f3ab55ccdbe7c57bcd21b292e485465fb566
Author: myui <yuin...@gmail.com>
Date: 2016-04-14T11:55:49Z
Changed resource location of unit tests
commit 3994cc7bb706e0b073278176eee6a1a5b78bebf6
Author: Sotaro Sugimoto <sotaro.sugim...@gmail.com>
Date: 2016-04-15T06:17:18Z
Both normal and PKI versions of KPA now learn identically.
Note that testTime was added to see how much faster PKI is, but it appears
that PKI is actually slower than normal KPA for non-sparse support vectors,
since PKI must go through the work of finding all matching support vectors,
which is O(number of support vectors) for a completely full match.
commit 31d981bd0082d4d5c40d7eb536f47f70455a3c52
Author: Sotaro Sugimoto <sotaro.sugim...@gmail.com>
Date: 2016-04-15T09:38:49Z
Fixed incorrect alpha calculation/storage.
Also fixed tests and asserts to match correct alpha calculations.
commit e2812fd035f51877167f0c77aff397fa45be0bef
Author: Sotaro Sugimoto <sotaro.sugim...@gmail.com>
Date: 2016-04-15T09:59:16Z
Added FloatArrayList.
Also changed alpha from DoubleArrayList to FloatArrayList.
commit c793d47e0de18d32a56d2d1fd1744470f7283613
Author: Sotaro Sugimoto <sotaro.sugim...@gmail.com>
Date: 2016-04-20T03:14:12Z
Added kernel expansion option (untested)
commit 4716b34ad5038761d18e422833d9334140dd0908
Author: Sotaro Sugimoto <sotaro.sugim...@gmail.com>
Date: 2016-04-21T03:34:26Z
Tests for KPA added.
Note that kernel splitting (in KernelExpansionKPA) is incomplete.
commit f986803901824ad831c7c8a2504c26797d765f23
Author: Sotaro Sugimoto <sotaro.sugim...@gmail.com>
Date: 2016-04-26T04:21:22Z
Kernel Expansion complete with tests.
Also removed Kernel Splitting (probably uses too much memory).
commit b0a0179b0bc1f50403eb1f5534bfb870113f9777
Author: myui <yuin...@gmail.com>
Date: 2016-12-02T07:55:49Z
Merge branch 'feature/kernelized_pa' of https://github.com/L3Sota/hivemall
into JIRA-22/pr-304
----
> Review and merge pending Pull Requests before entering Incubator
> ----------------------------------------------------------------
>
> Key: HIVEMALL-22
> URL: https://issues.apache.org/jira/browse/HIVEMALL-22
> Project: Hivemall
> Issue Type: New Feature
> Reporter: Makoto Yui
> Assignee: Makoto Yui
>
> Need to review and merge pending Pull Requests in
> https://github.com/myui/hivemall/pulls
> * Feature Selection
> https://github.com/myui/hivemall/pull/385
> * SST change point detection
> https://github.com/myui/hivemall/pull/356
> * Checkstyle
> https://github.com/myui/hivemall/pull/343
> * System Test
> https://github.com/myui/hivemall/pull/336
> * Kernelized Passive Aggressive
> https://github.com/myui/hivemall/pull/304
> * Separate Optimizer
> https://github.com/myui/hivemall/pull/285
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
