tqchen
|
98e507451c
|
chg all settings to obj
|
2014-05-16 19:10:52 -07:00 |
|
tqchen
|
213375baca
|
pre-release version
|
2014-05-16 18:49:02 -07:00 |
|
tqchen
|
2baeeabac4
|
new lambda rank interface
|
2014-05-16 00:02:26 -07:00 |
|
tqchen
|
6af6d64f0b
|
a correct version
|
2014-05-15 21:11:46 -07:00 |
|
tqchen
|
c22df2b31a
|
ok
|
2014-05-15 18:56:28 -07:00 |
|
tqchen
|
e2d13db24e
|
bug fix in pairwise rank
|
2014-05-15 15:37:58 -07:00 |
|
tqchen
|
37e1473cea
|
cleanup code
|
2014-05-15 15:01:41 -07:00 |
|
tqchen
|
a59969cd52
|
change data format to include weight in binary file, add get weight to python
|
2014-05-15 14:37:56 -07:00 |
|
tqchen
|
88526668f5
|
add ams
|
2014-05-14 23:23:27 -07:00 |
|
tqchen
|
31a0823e6d
|
some fix
|
2014-05-14 16:55:59 -07:00 |
|
tqchen
|
ae9d937510
|
add AMS metric
|
2014-05-14 11:30:45 -07:00 |
|
kalenhaha
|
121348c0d7
|
add in grad and hess rescale in lambdarank
|
2014-05-14 23:13:27 +08:00 |
|
kalenhaha
|
671c34be63
|
small bug in ndcg eval
|
2014-05-13 14:30:42 +08:00 |
|
kalenhaha
|
e858523d19
|
remove sampler
|
2014-05-11 14:31:57 +08:00 |
|
kalenhaha
|
faf35c409e
|
small change
|
2014-05-11 14:03:21 +08:00 |
|
tqchen
|
604568b512
|
simple chgs
|
2014-05-09 20:39:15 -07:00 |
|
kalenhaha
|
f7b2281510
|
fix some warnings
|
2014-05-09 14:14:43 +08:00 |
|
kalenhaha
|
0794dd0f6f
|
Merge branch 'dev' of https://github.com/tqchen/xgboost into dev
|
2014-05-09 14:07:06 +08:00 |
|
kalenhaha
|
4b6024c563
|
Separating Lambda MAP and Lambda NDCG
|
2014-05-09 14:05:52 +08:00 |
|
tqchen
|
a0c0fbbb61
|
commit the fix
|
2014-05-08 19:31:32 -07:00 |
|
tqchen
|
06327ff8d0
|
Merge branch 'dev' of ssh://github.com/tqchen/xgboost into dev
|
2014-05-07 12:00:17 -07:00 |
|
tqchen
|
0bf6261961
|
fix omp for bug in obj
|
2014-05-07 11:52:12 -07:00 |
|
kalenhaha
|
8b3fc78999
|
Merge branch 'dev' of https://github.com/tqchen/xgboost into dev
Conflicts:
regrank/xgboost_regrank_obj.hpp
|
2014-05-07 22:15:59 +08:00 |
|
tqchen
|
833cf29867
|
fix
|
2014-05-06 16:53:37 -07:00 |
|
tqchen
|
4b00b3e565
|
Merge branch 'dev' of ssh://github.com/tqchen/xgboost into dev
|
2014-05-06 16:51:18 -07:00 |
|
tqchen
|
abe5309977
|
Merge branch 'dev' of ssh://github.com/tqchen/xgboost into dev
Conflicts:
regrank/xgboost_regrank_data.h
|
2014-05-06 16:51:11 -07:00 |
|
tqchen
|
7ddff7b570
|
add regrank utils
|
2014-05-06 16:50:46 -07:00 |
|
tqchen
|
c39e1f2f30
|
right group size
|
2014-05-06 16:49:10 -07:00 |
|
tqchen
|
9bc699fd0e
|
add cutomized training
|
2014-05-04 13:55:58 -07:00 |
|
tqchen
|
8c0c10463e
|
add boost group support to xgboost. now have beta multi-class classification
|
2014-05-04 12:10:03 -07:00 |
|
kalenhaha
|
8eae8d956d
|
c++11 features removed
|
2014-05-04 16:58:44 +08:00 |
|
kalenhaha
|
7161618b4c
|
c++11 features removed
|
2014-05-04 16:56:57 +08:00 |
|
tqchen
|
6fd77cbb24
|
add python interface for xgboost
|
2014-05-03 23:04:02 -07:00 |
|
tqchen
|
adc9400736
|
finish python lib
|
2014-05-03 22:18:25 -07:00 |
|
tqchen
|
aab1b0e7b3
|
important change to regrank interface, need some more test
|
2014-05-03 14:20:27 -07:00 |
|
tqchen
|
cc91c73160
|
add new combine tool as promised
|
2014-05-02 12:55:34 -07:00 |
|
tqchen
|
ef7df40bc8
|
cleanup of evaluation metric, move c++11 codes into sample.h for backup, add lambda in a clean way latter
|
2014-05-01 11:00:50 -07:00 |
|
kalenhaha
|
f17d400fd3
|
fix some bugs in linux
|
2014-05-02 00:16:12 +08:00 |
|
kalenhaha
|
b836b1123e
|
lambda rank added
|
2014-05-01 22:17:26 +08:00 |
|
tqchen
|
bf64608cc9
|
add softmax
|
2014-04-30 22:11:26 -07:00 |
|
tqchen
|
54c482ffd5
|
add pre @ n
|
2014-04-30 22:00:53 -07:00 |
|
tqchen
|
bb93c0aaac
|
add rank
|
2014-04-30 09:32:42 -07:00 |
|
tqchen
|
a383f11759
|
add pairwise rank first version
|
2014-04-29 21:12:30 -07:00 |
|
tqchen
|
81414c0e5b
|
new AUC code
|
2014-04-29 17:26:58 -07:00 |
|
tqchen
|
87a9c22795
|
new AUC evaluator, now compatible with weighted loss
|
2014-04-29 17:03:34 -07:00 |
|
tqchen
|
31edfda03c
|
make regression module compatible with rank loss, now support weighted loss
|
2014-04-29 16:16:02 -07:00 |
|