Home

mlpack (machine learning pack) is an open-source, intuitive, fast, and scalable C++ machine learning library, meant to be a machine learning analog to LAPACK. It aims to implement a wide array of machine learning methods and function as a "swiss army knife" for machine learning researchers.

This wiki is meant to provide documentation for development-related tasks on Github.

mlpack depends on the following libraries:

• LAPACK
• BLAS
• Armadillo >= 3.6.0
• LibXml2 >= 2.6.0
• Boost >= 1.49 - program_options, math_c99, unit_test_framework, random, heap

All of these should be available in your distribution's package manager.

developer reference

• UsingCMake -- Tutorial for using/modifying our CMake configuration.
• WindowsBuild -- Tutorial for compiling mlpack on Windows.
• NewStyleGuidelines -- Coding style guidelines followed in mlpack code.
• The Future of mlpack -- mlpack design goals document (from 2010).
• Armadillo -- Project website for Armadillo; includes documentation.
• Jenkins -- mlpack build server.
• CompetingLibraries -- Information about competing libraries.

misc. reference

• ReleasePreparationGuide -- List of tasks to be done when releasing MLPACK.
• JenkinsSlaveConfiguration -- How to set up a slave for the build server.
• SummerOfCodeIdeas -- List of ideas for Google Summer of Code 2015.
• JenkinsUsers -- Notes on Jenkins user configuration.
• AutomaticBenchmark -- Notes regarding the automatic benchmark system.