|
Abstract:
|
As a result of the vast advancements in technology, we frequently come across data in high dimensions. The literature is rich with proposed Bayesian methods to dictate inference when the curse of dimensionality is present. In this paper, we propose to extend the utility of the Gaussian and Diffused-gamma (GD) prior for feature extraction when dealing with high-dimensional spatiotemporal data. A case study is presented with multisubject electroencephalography (EEG) data to identify active regions of the brain in the presence of stimuli. One goal of EEG analysis is to extract information from the brain in a spatiotemporal pattern and analyze the functional connectivity between different areas of the brain as a response to a certain stimulus. Using this approach, our application domain is predicting the risk of early-onset alcoholism. Performance of the GD prior under this paradigm is compared with existing methods used in literature. Although our method is applied to EEG measurements to examine effects of chronic exposure to alcohol on the brain, we may apply our method to different domains such as MRI measurements to analyze early-onset of mental-related illnesses such as dementia.
|
