This package implements a few Bayesian inference algorithms for von Mises Fisher based mixture models
- MvMF_inference.m: a Gibbs sampler for finite mixture of vMFs
- iMvMF_inference.m: a Gibbs sampler for infinite mixture or Dirichlet Process mixture of vMFs (Chinese restaurant process representation)
- Collapsed_iMvMF_inference.m: a Gibbs sampler for infinite mixture or Dirichlet Process mixture of vMFs with the mean vectors integrated
- Collapsed_iMCIvMF_inference.m: a Gibbs sampler for infinite mixture of conditionally independent vMFs with the mean vectors integrated
To get started, see demo.mlx
I will rewrite the whole program and upload it soon ...
Copyright (C) 2018 Lei Fang, lf28 at st-andrews.ac.uk; Free to distribute, modify, adapt etc.
Please cite the following paper if needed (also for more details) L. Fang, J. Ye and S. Dobson, "Sensor-Based Human Activity Mining Using Dirichlet Process Mixtures of Directional Statistical Models," 2019 IEEE International Conference on Data Science and Advanced Analytics (DSAA), Washington, DC, USA, 2019, pp. 154-163.