edhsmm 0.1.2

An(other) implementation of Explicit Duration HMM/HSMM in Python 3

edhsmm

An(other) implementation of Explicit Duration Hidden Semi-Markov Model (EDHSMM) in Python 3.

The EM algorithm is based on Yu (2010), while the Viterbi algorithm is based on Benouareth et al. (2008).

The code style is inspired from hmmlearn and jvkersch/hsmmlearn.

Note: This is the "main" branch, which is for release. Use the "test" branch for prototyping.

Implemented so far

• EM algorithm (with & without right-censoring)
• Scoring (log-likelihood of observation under the model)
• Viterbi algorithm
• Generate samples
• Support for multivariate Gaussian emissions
• Support for multiple observation sequences
• Support for multinomial (discrete) emissions

See the README in the test branch for the full timeline.

Dependencies

• python >= 3.5
• numpy >= 1.10
• scikit-learn >= 0.16
• scipy >= 0.19

Installation & Tutorial

pip install edhsmm

For tutorial, see the notebooks.

Found a bug? Suggest a feature? Please post on issues.