An Introduction to Bayesian Nonparametric Methods

DATES: 5-6 February 2020 (2 sessions)
TIME: 9:00 - 12:00 (a total of 6 hours)
LOCATION: BCAM Seminar room

ABSTRACT:
Bayesian methods play a key role in modern statistical modelling. The main aim of Bayesian nonparametric methods is to avoid dependence on critical parametric assumptions, thus robustifying parametric models, and also to provide a sensitivity analysis for such models by embedding them in a broader nonparametric model. This course provides an introduction to Bayesian nonparametric methods, and particular emphasis will be placed on models based on Dirichlet processes and Polya trees, with a view towards applications and software implementation. Special attention will be given to density estimation, regression, hierarchical models, and survival analysis.

PROGRAMME:
1. Review of Bayesian parametric statistics.
2. Bayesian computational tools.
3. Models for density estimation based on Dirichlet process mixtures and mixtures of finite Polya trees.
4. Models for density regression based on dependent Dirichlet process mixtures and dependent tail free processes.
5. Applications and software implementation.
6. Other Bayesian nonparametric models (e.g., Bernstein polynomials, Gaussian processes).


*Registration is free, but mandatory before January 31st. To sign-up go to https://forms.gle/2E6ScvJv83T3ekSX8 and fill the registration form.

Student grants are available. Please, let us know if you need support for travel and accommodation expenses in the previous form before January 10th.

• BCAM_Course_2020_02_05