Practical Machine Learning Lecture: Nonparametric Bayesian methods (Dirichlet processes)

Nonparametric Bayesian methods (Dirichlet processes)
Lecturer: Kurt Miller
Date: Apr 22

References

There are numerous references on Bayesian methods and Markov Chain Monte Carlo (MCMC) techniques. Two useful textbooks are:

Bayesian Data Analaysis. Gelman, Carlin, Stern, Rubin.
Monte Carlo Statistical Methods. Robert, Casella

Unfortunately, there currently are no good texts on the Dirichlet Process. However, a great set of references can be found in the references section of an upcoming workshop on nonparametric Bayesian methods. This list includes many of the foundational papers about Dirichlet Processes as well as references to many recent applications.

Tutorials on nonparametric Bayesian methods

Structured Bayesian nonparametric models with variational inference: friendly introduction to Bayesian ideas and Dirichlet processes from the stick-breaking point of view; was given to an natural language processing audience, so all of it deals with discrete data (Dirichlets and multinomials)
Dirichlet processes, Chinese restaurant processes and all that: presents many views of the Dirichlet process and many more advanced topics