Welcome to neuroHMM’s documentation!

The neuroHMM project started as a student project in collaboration with the Neuroscience Institute of La Timone. The aim of the project is to provide Machine Learning tools, initially adapted to neurophysiological signals, which allow spatial, temporal and frequency analysis of experimental data. Most of the tools developed are based on hidden Markov models (HMMs), and are inspired by the HMM-MAR library developed in Matlab by a team of Oxford researchers. Our library aims to be an extension of the scikit-learn library, a reference for Machine Learning experts, and follows as much as possible the scikit-learn Estimators development standards. Some of the tools developed are also based on the hmmlearn library, which implements the more classical HMMs with scikit-learn development standards.


On this website, you will find:

  • A presentation of the theory behind each of the implemented tools by following the Theory tab;

  • A guide to installing and using the library on the User Guide page;

  • Documentation of each class and associated methods in the API Reference;

  • Example notebooks of how to use the Estimators on the Examples page.

Indices and tables