Commits · 9d375069f1536d97a67ba8d57ac75ca9e62f3341 · tianlh / LightGBM-DCU

13 Mar, 2017 2 commits
- fix max_bin of categorical feature in parallel learning. · 12ce2566
  Guolin Ke authored Mar 13, 2017
  
  12ce2566
- fix bugs for parallel learning. · ebc0de8b
  Guolin Ke authored Mar 13, 2017
  
  ebc0de8b
07 Mar, 2017 1 commit
- fix a bug in multi-class · 07f709b9
  Guolin Ke authored Mar 07, 2017
  
  07f709b9
06 Mar, 2017 1 commit
- fix typos (#336) · fff84463
  zhangyafeikimi authored Mar 06, 2017
  
  fff84463
05 Mar, 2017 1 commit
- fix comments (#334) · 45960910
  tks authored Mar 05, 2017
  
  45960910
03 Mar, 2017 1 commit
- fix one bug in 4bits_bin . · bc9d57b3
  Guolin Ke authored Mar 03, 2017
  
  bc9d57b3
01 Mar, 2017 12 commits
- Add categorical feature support back. · ef778069
  Guolin Ke authored Mar 01, 2017
  
  ef778069
- fix typos and comments (#326) · 8e190f59
  zhangyafeikimi authored Feb 27, 2017
  
  8e190f59
- slight speed up the iterator of sparse bin. · 362f9ac4
  Guolin Ke authored Feb 28, 2017
  
  362f9ac4
- fix index out of range bug. · 58000683
  Guolin Ke authored Feb 27, 2017
  
  58000683
- speed up subset of 4bits_bin. · 5e3b7193
  Guolin Ke authored Feb 27, 2017
  
  5e3b7193
- fix subset bug in 4bits_bin. · fd5b5916
  Guolin Ke authored Feb 27, 2017
  
  fd5b5916
- unrolling loop of histogram buliding in order_bin. · a05e8955
  Guolin Ke authored Feb 26, 2017
  
  a05e8955
- output shrinkage for trees (#320) · 4228bb78
  Guolin Ke authored Feb 26, 2017
  
  4228bb78
- Squashed commit of the following: · 9ea487b1
  Guolin Ke authored Feb 22, 2017
```
commit 70f88c54f6820d4c3824d97b34042c617fb21635
Author: Guolin Ke <i@yumumu.me>
Date:   Wed Feb 22 19:40:10 2017 +0800

    futher reduce guided

commit f54620b807c68f5ac6cef521d787ca2750efe464
Author: Guolin Ke <i@yumumu.me>
Date:   Wed Feb 22 19:32:36 2017 +0800

    avoid to use guided
```
  9ea487b1
- add 4bits_bin · a8bb3951
  Guolin Ke authored Feb 22, 2017
  
  a8bb3951
- tab to space, improve comments typos (#302) · 8b00b8e1
  zhangyafeikimi authored Feb 20, 2017
  
  8b00b8e1
- update to v2. · 4f77bd28
  Guolin Ke authored Feb 20, 2017
  
  4f77bd28
30 Jan, 2017 1 commit
- Support Poisson regression (#270) · 4f3e9d89
  Tsukasa OMOTO authored Jan 31, 2017
```
* Support Poisson regression

Close #233

* update docs
```
  4f3e9d89
26 Jan, 2017 2 commits
- fix warning · da59fd22
  Guolin Ke authored Jan 26, 2017
  
  da59fd22
- not limit max_bin for categorical feature · 68bc8a61
  Guolin Ke authored Jan 26, 2017
  
  68bc8a61
25 Jan, 2017 2 commits
- support negative values for sparse features. · 9962e6d6
  Guolin Ke authored Jan 25, 2017
  
  9962e6d6
- fix partition error when set weight_colunm · 1765b2e3
  Guolin Ke authored Jan 25, 2017
  
  1765b2e3
24 Jan, 2017 2 commits
- fix bug for subset · 6a367ac4
  Guolin Ke authored Jan 24, 2017
  
  6a367ac4
- reduce function call cost for constructing subset. · 3489607f
  Guolin Ke authored Jan 24, 2017
  
  3489607f
23 Jan, 2017 3 commits
- [bug fix] fix predict sigmoid; fix bagging bug. · 9c5dbdde
  Guolin Ke authored Jan 23, 2017
  
  9c5dbdde
- slight improve prediction speed · fd0cbe65
  Guolin Ke authored Jan 23, 2017
  
  fd0cbe65
- improve bagging speed · 4948dcc5
  Guolin Ke authored Jan 23, 2017
  
  4948dcc5
22 Jan, 2017 1 commit
- use subset to speed up bagging · c8fbd42b
  Guolin Ke authored Jan 22, 2017
  
  c8fbd42b
14 Jan, 2017 2 commits
- not check metrics w/o eval set (#212) · 2e100eeb
  wxchan authored Jan 14, 2017
  
  2e100eeb
- output more feature information in model file, #192 · 011fe024
  Guolin Ke authored Jan 14, 2017
  
  011fe024
12 Jan, 2017 1 commit

Make openmp optional, allow for dylib mac library output (#204) · 300cee50

cbecker authored Jan 12, 2017

* Added possibility to compile without openmp (useful for clang on Mac)

* Added CMake option to create dylib on MacOS

300cee50

10 Jan, 2017 5 commits
- fix bug in sparse bin iterator · c2ba086c
  Guolin Ke authored Jan 10, 2017
  
  c2ba086c
- change init_score to double type · 3ef3a489
  Guolin Ke authored Jan 10, 2017
  
  3ef3a489
- change inner prediction score to double type. · 12a96334
  Guolin Ke authored Jan 10, 2017
  
  12a96334
- fix name of boosting type · dd316895
  Guolin Ke authored Jan 10, 2017
  
  dd316895
- Add a parameter to control width of Gaussian function to approximate hessian (#182) · 26027d35
  Tsukasa OMOTO authored Jan 10, 2017
  
  26027d35
09 Jan, 2017 3 commits
- Add Fair loss (#180) · fb732c34
  Tsukasa OMOTO authored Jan 10, 2017
```
http://research.microsoft.com/en-us/um/people/zhang/INRIA/Publis/Tutorial-Estim/node24.html
```
  fb732c34
- use std::string for tree_learner_type. · a6f47d00
  Guolin Ke authored Jan 09, 2017
  
  a6f47d00
- Add Huber loss (#174) · a87af879
  Tsukasa OMOTO authored Jan 09, 2017
```
* Add Huber loss

https://en.wikipedia.org/wiki/Huber_loss

* update

* update docs/Parameters.md

* update
```
  a87af879