Algorithm Library Architecture

The machine learning algorithm library provides the same APIs as Spark MLlib. The library consists of the core algorithm implementation packages of Boostkit-ML-Kernel and the ML-API-Patch code for machine learning that interconnects with native Spark APIs.

As shown in the preceding figure, ML-API-Patch corresponds to boostkit-ml-core_2.11-1.3.0-spark2.3.2.jar, boostkit-ml-acc_2.11-1.3.0-spark2.3.2.jar, boostkit-xgboost4j_2.11-1.3.0.jar, boostkit-xgboost4j-spark2.3.2_2.11-1.3.0.jar compiled by using the open source patch of the library. Boostkit-ML-Kernel corresponds to the packages of the machine learning algorithm library boostkit-ml-kernel_2.11-1.3.0-spark2.3.2-aarch64.jar, boostkit-xgboost4j-kernel-2.11-1.3.0-spark2.3.2-aarch64.jar, and libboostkit_xgboost_kernel.so.

This document uses the BoostKit algorithm package based on Spark 2.3.2 as an example and also applies to the BoostKit algorithm package based on Spark 2.4.6.