Title: Bayesian Ising sparse nonparametric model
Authors: Inyoung Kim - Virginia Tech (United States) [presenting]
Abstract: A Bayesian variable selection approach is proposed via the graphical model and the Ising model. Our Bayesian variable problem can be considered as a complete graph and described by an Ising model with random interactions. There are several advantages of our approach: it is easy to (1) employ the single-site updating and cluster updating algorithm, both of which are suitable for problems with small sample sizes and a larger number of variables, (2) extend this approach to other regression models, and (3) incorporate graphical prior information. In our approach, the interactions are determined by the linear model coefficients, so we systematically study the performance of different scale normal mixture priors for the model coefficients by adopting the global-local shrinkage strategy. Our results indicate that the best prior for the model coefficients in terms of variable selection should place substantial weight on small, nonzero shrinkage. The methods are illustrated with simulated and pathway genomics data.