Differential Modeling for Biomedical Applications
Dr. Chandan Reddy
Computer Science Department
Wayne State University
3:45 pm, March 7, 2011, Room 228 MSB
In many of the practical applications, data analysts are typically interested in finding some specific patterns that distinguish two subgroups within complex multivariate datasets. Detecting and describing the discriminations across multiple multivariate datasets is a challenging problem that arises in many areas of science and engineering. More formally, the primary objective of this differential modeling task is to quantify, summarize, and build models that capture the discriminations between the distributions of the subgroups in the data and identify discriminative patterns that are specific to a particular subgroup. In this talk, I will describe several algorithms and approaches that we proposed to analyze discriminations across multiple subgroups through predictive patterns, network modules and subspace correlations that are distinct to specific subgroups in complex datasets. Our methods provide effective summaries of the discriminations across multiple groupings of the data and allow the domain experts to gain more understanding about the behaviour of these subgroups in the data. Our work was rigorously evaluated using important biological and epidemiological studies and our experimental results showed that the patterns obtained are both statistically and biologically significant. We applied our differential modeling algorithms to biomedical datasets containing different racial groups (Caucasian-American and African-American) and different progressive stages of cancer (tumor and normal nonmalignant).
Chandan Reddy is an Assistant Professor in the Department of Computer Science at Wayne State University since Fall 2007. He received his PhD from Cornell University and MS from Michigan State University. He is the Director of the COmputational Learning and Discovery (COLD) Laboratory and a scientific member of Karmanos Cancer Institute. His primary research interests are machine learning and data mining with applications to biomedical informatics and business intelligence. He has published over 30 peer-reviewed articles in leading conferences and journals including IEEE TPAMI, ACM SIGKDD, IEEE ICDM, SIAM DM, and ACM CIKM. He received the Best Application Paper Award in ACM SIGKDD 2010, and the Faculty Research Award from WSU in 2008. He is a member of IEEE and ACM.