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
However, in our test on the ranking data-set, we didn't manage to make lambdaRank output-perform pairwise ones, this makes us hold the release of current code.
Yes, the rank module is designed to be lambdaRank, see the incoming new branch https://github.com/tqchen/xgboost/tree/unity https://github.com/tqchen/xgboost/blob/unity/src/learner/objective.h
However, in our test on the ranking data-set, we didn't manage to make lambdaRank output-perform pairwise ones, this makes us hold the release of current code.