Student & Alumni Achievements

Student & Alumni Achievements

UAB PhD Candidate Wins ASA Award in Student Paper Competition

“The Bayesian Group Bridge,” a paper by Mr. Himel Mallick, PhD candidate in the department of biostatistics, section on statistical genetics (SSG), at the University of Alabama at Birmingham — co-authored with Dr. Nengjun Yi, Sir David Cox Endowed Professor in Biostatistics at UAB — has been selected by the American Statistical Association (ASA) as one of the 10 best papers submitted in their Section on Bayesian Statistical Science (SBSS) Student Paper Competition. Mr. Mallick will present a talk on the topic of the winning paper at the JSM (Joint Statistical Meetings) 2015, to be held August 8-13, in Seattle, Washington.

[Photo: Mr. Himel Mallick]

The paper deals with variable selection for grouped data. In many scientific applications, covariates are naturally grouped. One such example arises in genetic association studies, where genetic variants can be divided into multiple groups, with variants in a same group being biologically related or statistically correlated. In such situations, traditional variable selection methods (such as the LASSO and related methods) may often lead to poor performance, as they usually do not take into account the grouping information into their estimation procedures. The proposed method overcomes this limitation by taking into account the relevant grouping information into its estimation procedure. Empirical evidence of the attractiveness of the method is illustrated by extensive simulations and real data analyzes.

From August 2010 to August 2014, Mr. Mallick was a graduate research assistant in the REasons for Geographic and Racial Differences in Stroke (REGARDS) study. He is currently serving as a graduate research assistant under the supervision of Dr. Hemant Tiwari, head of UAB’s SSG. Mr. Mallick’s research interests lie in the application of modern statistical and machine learning tools to high-dimensional genomic data and the development of novel Bayesian methods for analyzing genetics and clinical trial datasets.