Bayesian Hidden Markov Model with Gibbs sampling
Python 2.7.3+
./cluster.py <input file> [output file] <number of labels> <number of iterations> <alpha> <beta>
<input file> - File with sequences, one per line
[output file] - Output file
<number of labels> - Possible labels
<number of iterations> - Number of sampling iterations
<alpha> - Hyperparameter for transitions
<beta> - Hyperparameter for emissions
- bhmm.py - Gibbs sampling
- cluster.py - Entry point
- utility.py - Utility methods
- corpus.txt - toy corpus
MIT