Releases: kaselby/DirichletMixtureModels.jl
Releases · kaselby/DirichletMixtureModels.jl
First Experimental Version
This is a first experimental version of my package to perform clustering over mixture models with a Dirichlet Process prior. It currently supports the conjugate distributions Exponential/Gamma, Normal/Normal-Gamma, Normal (Known Sigma)/Normal, Multivariate-Normal/Normal-Wishart, as well as user-defined distributions (conjugate or non-conjugate). Clustering is performed using Neal's algorithms 2 and 8 (see the resources section).
Currently, the package overrides several methods from the ConjugatePriors package due to errors with the NormalWishart implementation. This will be updated once these are fixed.