Lead maintainer of XGBoost. Area of interest: Machine learning systems & algorithms.
-
NVIDIA
- Los Angeles, CA, USA
- https://hyunsu-cho.io
- @hcho3_ml
Highlights
- Pro
Block or Report
Block or report hcho3
Report abuse
Contact GitHub support about this user’s behavior. Learn more about reporting abuse.
Report abusePinned
-
dmlc/xgboost Public
Scalable, Portable and Distributed Gradient Boosting (GBDT, GBRT or GBM) Library, for Python, R, Java, Scala, C++ and more. Runs on single machine, Hadoop, Spark, Dask, Flink and DataFlow
-
-
694 contributions in the last year
Activity overview
Contribution activity
May 2022
Created 12 commits in 5 repositories
Created a pull request in rapidsai/cuml that received 6 comments
Upgrade Treelite to 2.4.0
The 2.4.0 version of Treelite incorporates the following improvements: dmlc/treelite#370 dmlc/treelite#381 dmlc/treelite#380 dmlc/treelite#383 Re…
+9
−9
•
6
comments
Opened 10 other pull requests in 4 repositories
dmlc/treelite
4
merged
dmlc/xgboost
4
merged
rapidsai/integration
1
merged
conda-forge/treelite-feedstock
1
merged
Reviewed 16 pull requests in 2 repositories
dmlc/xgboost
15 pull requests
- Verify shared object version at load.
- Refactor for GHistIndex.
- Add convergence test for sparse datasets.
- Simplify inplace-predict.
- Use adapter to initialize column matrix.
- Remove tree structure assumption in GetNextNode
- [Breaking][jvm-packages] make classification model be xgboost-compatible
- Extract partial sum into an independent function.
- [breaking] Drop single precision hist.
- Fix monotone constraint with tuple input.
- Show libxgboost.so path in build info.
- Fix overflow in prediction size.
- Update CUDA version requirement in CMake script.
- Allow loading old models from RDS
- Use maximum category in sketch.
dmlc/treelite
1 pull request
Created an issue in dmlc/treelite that received 6 comments
Support XGBoost 1.6
Handle the extra slope parameter in the pseudo-Huber loss: #380 Support UBJSON serialization format Support multi-output regression. Support m…
2 of 4 tasks