mlpack
a scalable c++ machine learning library

mlpack benchmarks


  This page contains benchmarks for the various algorithms implemented in mlpack 1.0.10. When applicable, timing results are also given for other libraries. Currently, benchmarks are provided for scikit-learn, the SHOGUN machine learning toolbox, Weka 3, mlpy, and standard MATLAB implementations. If you don't see benchmarks for your favorite library or algorithm, feel free to file a bug report or consult the documentation for the automatic benchmarking system and write a script.

  The automatic benchmarking system was developed as a Google Summer of Code 2013 project by Marcus Edel and improved by Anand Soni during GSoC 2014, and is maintained separately on Github.

  Below, you may select one of many interactive JavaScript visualizations for inspecting the benchmarking results. At this time, we are still adding results (especially historical results). If you are interested in the previous non-interactive visualizations, which included memory usage graphs for mlpack methods, click here.