Speaker Bio: Sumit K. Jha is an Assistant Professor with the Computer Science department at University of Central Florida. He graduated from Christopher Langmead's research group at Carnegie Mellon with a PhD in Computer Science. Before joining Carnegie Mellon, Sumit graduated with B.Tech (Hons.) in Computer Science from the Indian Institute of Technology Kharagpur.
His current research interests include verification and synthesis of stochastic and hybrid systems with emphasis on applications to biochemical modeling and computational finance. Sumit has also worked on more traditional formal verification and synthesis problems with Microsoft Research and General Motors.
Title of the Talk: Parameter Synthesis for Stochastic Biochemical Models from High Level Behavioral Specifications
Abstract: Traditional algorithms for parameter synthesis often use time series experimental data to estimate a single most likely parameter value for the kinetic rate constants in stochastic biochemical models. Often, a manual hit and trial approach is also used to choose the kinetic rate constants so as to construct a model that closely reflects the behavior of the biochemical system under study. In this talk, we will suggest algorithms to synthesize all possible parameter values that enable a stochastic biochemical model to satisfy a given list of high level behavioral properties. As the models are stochastic, we permit the use of probabilistic behavioral specifications. Our solution is based on a combination of statistical verification of the biochemical stochastic model against behavioral specifications, and a systematic exhaustive exploration of the parameter space. We will discuss the benefits of such a synthesis approach, the ongoing applications of our algorithms and then present opportunities for future research.