NumFOCUS is delighted to announce the addition of the Shogun Machine Learning Toolbox to our fiscally sponsored projects.
Shogun’s mission is to make powerful machine learning tools available to everyone —researchers, engineers, students — anyone curious to experiment with machine learning to leverage data. The Shogun Machine Learning Toolbox provides efficient implementations of standard and state-of-the-art machine learning algorithms in an accessible, open-source environment.
A core strength of Shogun is that its internals, written in modern C++, can be interfaced from many languages, including Python, Octave, R and more, under a unified interface. This approach allows for rapid prototyping in high-level languages while harnessing the benefits of efficient low-level implementations. See the examples at shogun.ml/examples or try a free cloud version of the Python interface at cloud.shogun.ml
Shogun was initiated in 1999 by researchers Soeren Sonnenburg and Gunnar Raetsch, who at the time pushed the boundaries of large-scale multiple kernel methods. Since then, Shogun has evolved into a general toolbox covering many machine learning algorithms as well as general purpose methods such as for evaluation and parameter tuning. This development has been particularly driven by the support from Google Summer of Code, which Shogun has participated in since 2011.Many contributors who got involved with Shogun as students are now successful researchers and software engineers, and all acknowledge Shogun as a fundamental pillar in their education. Extending its impact in machine learning education beyond individuals, Shogun is currently being used in courses taught at University College, London, UK.
The Shogun leadership body for NumFOCUS consists of Gunnar Rätsch, Sören Sonnenburg, Viktor Gal, Fernando Iglesias, Sergey Lisitsyn, and Heiko Strathmann.