A collection of low-level machine learning algorithms for node.js.
This project is quite new and documentation will be on the way shortly. In the meantime you can check out the spec folder for examples of how to use the algorithms.
Note that within "apparatus" the interface to the algorithms in primarily arrays of numbers and vectors. If you're looking for feature extraction from text or natural language check out the "natural" https://github.com/NaturalNode/natural node package. "natural" uses many of these algorithms but adds a layer of natural language/text feature extraction.