Implements the infinite Hidden Markov Model for diverse single-molecule experiments.
This project contains Matlab scripts that implement the infinite Hidden Markov Model (iHMM) [1, 2] for specific datasets from single-molecule experiments.
The code can be used out of the box for similar applications or serve as a guide to develop new emission models. Applying the scripts to one of the datasets provided would be a good tutorial on analyzing the output of non-parametric HMMs.
[1] Van Gael, J. et al. Beam sampling for the infinite Hidden Markov Model. Proceedings of the 25th International Conference of Machine Learning, 2008.
[2] Teh, YW. Hierarchical Dirichlet Process. Journal of the American Statistical Association. Vol. 101, No. 476, p. 1566, 2006.