Postdoctoral Fellow - Biostatistics

The Department of Biostatistics at UT MDACC has one post-doc position open for biostatistics methodology research with a focus on the analysis of neuroimaging data and integrative analysis of brain imaging and genetics data. The main focus is research and publication. The primary goals will be to develop novel methodology and computational algorithms for the identification and validation of predictive neuroimaging markers for early detection of mental disorders and their relation to -omics features, derive related statistical theory to justify the methodology, produce software for implementation, and build statistical learning tools for large data sets.

The candidate does not need prior expertise in neuroimaging studies. The post-doc will work under the supervision of Dr. Suprateek Kundu on challenging and important projects that involve complex and high-dimensional imaging data along with -omics features. There will be a scope of working on important and cutting-edge problems, including Bayesian joint estimation of multiple networks, tensor modeling that accounts for heterogeneity within tumors and across subjects, Bayesian non- and semi-parametric functional regression models, and so on.

LEARNING OBJECTIVES
Learn statistical methodology and theory for functional linear models, scalar-on-function imaging Tensor response regression, Bayesian non- and semi-parametric modeling of multivariate and temporally dependent data, statistical modeling for high-dimensional networks and probabilistic graphical models that are either stationary or dynamic over time.

Develop strong programming and computational skills in Bayesian computation; Gain interest in statistical methodology research; Adapt and develop novel machine learning algorithms and develop theoretical properties involving Bayesian posterior consistency and finite sample properties for sample estimators;

Acquire extensive experience in Matlab, R, R-Shiny, Python, and other programming languages/environments.

ELIGIBILITY REQUIREMENTS
Applicants must have a recent PhD in biostatistics/statistics/machine learning/computer science or within 0-1 years of graduation. At least one first author publication in a peer reviewed journal stemming from PhD studies and pre-prints for additional manuscripts related to the doctoral dissertation where possible. A solid background in statistical theory, Bayesian methods, variable selection, spatial-temporal statistics, and machine learning is required. Some experience with biomedical applications is desirable.

FACULTY MENTOR
Dr. Suprateek Kundu, Associate Professor