Instructor: Roee Gutman
Textbook: Gelman, A., Carlin, J.B., Stern, H.S. and Rubin, D.B. (2014). Bayesian Data Analysis (3rd Edition). Chapman and Hall/CRC. ISBN-13 978-1-439-84095-5.
Description: The Bayesian paradigm for data analysis can be described as consisting of four main steps: constructing probability models for data given parameters, computing posterior distributions of parameters, exploring posterior distributions, and checking/improving models. Topics covered ranged from basic Bayesian models to the more complicated hierarchical and mixture models, many of which have important applications in a variety of fields. This course sought to expose students to the Bayesian approach and its conceptual underpinnings, ultimately fostering an ability to create original computer code for advanced statistical models and techniques.
The Bayesian approach to statistics differs from the classical or “frequentist” approach in several ways, offering us the flexibility to incorporate prior knowledge and experiences into data analyses. Modern advances in computing have facilitated the analysis of complex models that are challenging to examine using traditional frequentist methods, making them more accessible for analysis through Bayesian methodology.
Assignments:
Spring ‘22