xadupre / machinelearning_xgboost

XGBoost Wrapper for Microsoft.ML - https://github.com/dotnet/machinelearning

XGBoost Wrapper for Microsoft.ML

This project proposes an extension to machinelearning written in C# which wraps XGBoost into the C# API. The wrapper requires changes to XGBoost source code to enable one-off predictions available in this repository: XGBoost.

Build

On windows: build.cmd.

On Linux: build.sh.

Build the documentation: doxygen conf.dox.

Usage

Once the project machinelearning_xgboost is built, the binaires XGBoost needs to be copied in the same folder as Microsoft.ML's. Example:

machinelearning_xgboost's Issues

implements entrypoints

add artifacts to CI

remove submodule machinelearning to use nuget package

update the list of parameters to the latest version of XGBoost

update to latest machinelearning and xgboost

add unit tests for other models (ranker, ...) and other sets of parameters (dart, ...)

add an example in README.md

add step to build nuget

entrypoints are missing