Mixture Models

Mixture models are used to infer properties about sub-populations when presented with pooled population data. Several applications such as topic modeling, clustering, and rich-prior distribtion selection can be handled with mixture models. DMixLearn offers a flexible implementation of the mixture model. DMixLearn’s mixture model API combined with base distributions and combinators.` allows for the specification of deep nested graphical models on heterogenous data types.

Mixture Models

• Mixture Distribution
• Heterogeneous Mixture Distribution
• Joint Mixture Distribution
• Hierarchical Mixture Distribution
• Hidden Markov Distribution