/tf_dpgmm

Variational inference in Dirichlet process Gaussian mixture model (tensorflow implementation)

Primary LanguagePython

tf_dpgmm

Variational inference in Dirichlet process Gaussian mixture model (tensorflow implementation), for spherical and diagonal covariance models

There is a folder for each model type, and each contains:

  • a pdf containing the equations and derivations for the evidence lower bound and variational updates
  • dpgmm_vi.py: a tensorflow implementation of variational inference in the model
  • bound_check.py: a comparison of the analytical and Monte Carlo estimates of the ELBO. We used this to check that our derivations and code are correct, because the two estimates match.
  • demos.py: examples of how to use dpgmm_vi.py, including plotting changes in ELBO with each update and clustering results

These codes have not been optimized for performance. Please let us know if you find any mistakes!