Dirichlet process mixture models for non-stationary data streams

Ioar Casado, Aritz Pérez

In recent years, we have seen a handful of work on inference algorithms over non-stationary data streams. Given their flexibility, Bayesian non-parametric models are a good candida…