Online Program

Return to main conference page
Saturday, May 19
Computational Statistics
Image and High-Dimensional Processing
Sat, May 19, 1:15 PM - 2:45 PM
Regency Ballroom A
 

Bayesian Variable Selection Using Spike and Slab Prior With Application to High Dimensional EEG Data by Local Modeling (304575)

Dipak Dey, University of Connecticut 
*Shariq Mohammed, University of Connecticut 
Yuping Zhang, University of Connecticut 

Keywords: local Bayesian modeling, multi-subject neuroimaging, parallel computing, scale mixture of normal link, spike and slab prior, tensor data

Due to the immense technological advances, very often we encounter data in high-dimensions. Any set of measurements taken at multiple time points for multiple subjects leads to data of more than two dimensions (matrix of covariates for each subject). In this paper, we present a Bayesian method for binary classification of subject-level responses by building binary regression models using latent variables along with the well known spike and slab priors. We incorporate the scaled normal priors on the parameters, as they cover a large family of distributions. Due to the computational complexity, we build many local (at different time points) models and make predictions using the temporal structure between the local models. We do variable selection for each of these local models. If the variables are locations, then the variable selection can be interpreted as spatial clustering. We show the results of a simulation study and also present the performance of these models on multi-subject neuroimaging (EEG) data to study the effects on the functional states of frontal cortex and parietal lobe for chronic exposure of alcohol.