Bayesian Feature Extraction using Gaussian and Diffused-gamma Priors for High Dimensional Spatio-Temporal Data

2026-07-27

AI Summary

This study addresses the challenge of accurately and interpretably selecting features from high-dimensional spatiotemporal data, which often exhibit sparse structures and complex dependencies that traditional methods struggle to capture. The authors propose a Bayesian feature extraction framework that innovatively combines Gaussian and spike-and-slab gamma priors to induce structured sparsity, coupled with a general likelihood model based on Bregman divergence. A two-stage posterior inference procedure is developed, integrating MCMC sampling, false discovery rate (FDR) control, and clustering analysis to achieve stable and efficient spatiotemporal feature selection. Applied to multi-subject EEG data, the method successfully identifies brain regions significantly associated with chronic alcohol exposure, substantially improving both the accuracy of sparse feature recovery and model interpretability.

Abstract

High-dimensional data with sparse structure and spatio-temporal dependence arise in many scientific domains. We develop a Bayesian feature-extraction framework for spatio-temporal settings that employs Gaussian and Diffused-gamma priors to induce structured sparsity. The modeling framework specifies a general likelihood via Bregman divergence, enabling compatibility with a range of loss functions and measurement models. Posterior computation is carried out via Markov chain Monte Carlo (MCMC), and we introduce a two-stage feature-extraction procedure based on posterior samples to stabilize selection across space and time. We illustrate the method with a multi-subject electroencephalography (EEG) case study examining the relationship between chronic alcohol exposure and activity in different brain regions. We first fit binary classification models at each time point, then use false discovery rate-controlled screening and subsequent clustering in a two-stage feature-extraction pipeline to identify active brain regions. The analysis demonstrates that our proposed priors improve recovery of sparse features and enhance interpretability in the presence of spatio-temporal dependence. The framework is broadly applicable to high-dimensional, structured problems where accurate feature selection and inference are required. The code to implement the model is publicly available via GitHub.

Problem

• high-dimensional data
• spatio-temporal dependence
• structured sparsity
• feature extraction
• Bayesian inference

Innovation

• Bayesian feature extraction
• Diffused-gamma prior
• spatio-temporal sparsity
• Bregman divergence
• two-stage selection