- Updated XGBoost to 2.1.1
- Added support for callbacks
- Added
num_features
andsave_config
methods toBooster
- Added
num_nonmissing
anddata_split_mode
methods toDMatrix
- Dropped support for Ruby < 3.1
- Updated XGBoost to 2.0.0
- Dropped support for Ruby < 3
- Fixed error with
dup
andclone
- Updated XGBoost to 1.7.5
- Added musl shared library for Linux
- Improved error message for invalid matrix
- Updated XGBoost to 1.7.0
- Updated XGBoost to 1.6.1
- Improved ARM detection
- Dropped support for Ruby < 2.7
- Updated XGBoost to 1.5.0
- Updated XGBoost to 1.4.0
- Added ARM shared library for Linux
- Added ARM shared library for Mac
- Fixed error with validation sets without early stopping
- Updated XGBoost to 1.3.0
- Updated XGBoost to 1.2.0
- Updated XGBoost to 1.1.0
- Changed default
learning_rate
andmax_depth
for Scikit-Learn API to match Python - Added support for Rover
- Improved performance of Numo datasets
- Improved error message when OpenMP not found on Mac
- Added
feature_names
andfeature_types
toDMatrix
- Added feature names to
dump
- Updated XGBoost to 1.0.0
- Fixed
Could not find XGBoost
error on some Linux platforms - Fixed
SignalException
on Windows
- Prefer
XGBoost
overXgb
- Changed to Apache 2.0 license to match XGBoost
- Added shared libraries
- Added support for booster attributes
- Added support for missing values
- Fixed Daru training and prediction
- Fixed error with JRuby
- Friendlier message when XGBoost not found
- Free memory when objects are destroyed
- Added
Ranker
- Added early stopping to Scikit-Learn API
- Added Scikit-Learn API
- Added early stopping
- Added
cv
method - Added support for Daru and Numo::NArray
- Added many other methods
- Fixed shape of multiclass predictions when loaded from file
- First release