Bayesian Statistics Seminar
North Carolina State University
presents
Dr. Sujay Datta
Northern Michigan University
"Some examples of applications of Bayesian methodology in bio-medical sciences"
ABSTRACT
Problems arising from the bio-medical sciences often present difficult
challenges to modern statistics---not only because traditional areas such
as epidemiology and clinical trials are having to deal with a much larger
and more ethnically/culturally interwoven human population with much
greater wordwide mobility and increasingly drug-resistant pathogens, but
also because rapidly advancing medical technology has brought about an
'information revolution' inundating statisticians with an unprecedented
amount of data. Age-old statistical methodology is often found struggling
to cope up with the new demands of modern-day medical science and statistics
as a discipline is going through a constant process of adaptation and
modification to prove itself capable of converting information into knowledge.
Aided by high-performance computing, the Bayesian approach together with
MCMC techniques has provided useful and practical solutions to many
seemingly intractable problems in bio-medical sciences. Here we discuss a
couple of them.
First, a Bayesian nonlinear mixed-effects model for the propagation
of the HIV virus in human body is discussed, together with its scientific
justification, analysis and possible improvements. Next, a Bayesian analysis
of gene locus ordering in the context of human chromosome breakage under
the radiation hybrid mapping technique is discussed.
Tuesday, November, 9, 2004
4:00 - 5:00 pm
208 Patterson Hall
Refreshments will be served on the second floor of Patterson Hall (outside Room 208) at 3:45 pm.