Philip Hyunsu Cho 44811f2330

Fix #3485, #3540: Don't use dropout for predicting test sets (#3556) ...

* Fix #3485, #3540: Don't use dropout for predicting test sets

Dropout (for DART) should only be used at training time.

* Add regression test

2018-08-05 10:17:21 -07:00

..

c_api

Fix #3545: XGDMatrixCreateFromCSCEx silently discards empty trailing rows (#3553)

2018-08-05 10:15:42 -07:00

common

Dynamically allocate GPU histogram memory (#3519)

2018-07-28 21:22:41 +12:00

data

Add qid like ranklib format (#2749)

2018-06-30 20:24:03 +00:00

gbm

Fix #3485, #3540: Don't use dropout for predicting test sets (#3556)

2018-08-05 10:17:21 -07:00

linear

Resolve GPU bug on large files (#3472)

2018-07-16 20:43:45 +12:00

metric

Clang-tidy static analysis (#3222)

2018-04-19 18:57:13 +12:00

objective

Fix tweedie handling of base_score (#3295)

2018-06-28 15:43:05 +00:00

predictor

Fix #3505: Prevent undefined behavior due to incorrectly sized base_margin (#3555)

2018-08-05 10:14:07 -07:00

tree

Fix #3402: wrong fid crashes distributed algorithm (#3535)

2018-08-04 19:20:04 -07:00

cli_main.cc

Save outputs in high precision in CLI prediction (#3356)

2018-06-03 14:15:47 -07:00

learner.cc

Fix #3485, #3540: Don't use dropout for predicting test sets (#3556)

2018-08-05 10:17:21 -07:00

logging.cc

Add callback interface to re-direct console output (#3438)

2018-07-05 11:32:30 -07:00