Global Dynamics in Neural and Social Systems: What is Tricky about Them?

Abstract: Complex systems such as the brain and social systems consist of many components that interact with each other nonlinearly. The control of such systems using local interventions often has unintended consequences system-wise. This is a general difficulty associated with high-dimensional nonlinear dynamical systems: it is easy to model the system near a specific state but hard to tell how it can go to other distinct states. In this talk, I will show why modeling neural and social dynamics requires a global approach. That is, rather than focusing on the mechanism that stabilizes a specific state of the brain or social interaction, one should look more at the global landscape of the dynamical systems: where are all the stable states? How can they transition from one to another? I will show how such a global dynamic landscape can be charted using dynamical systems modeling and topological data analysis and how it applies to basic and clinical neuroscience.

Bio: Dr. Mengsen Zhang is an assistant professor at Michigan State University in the Department of Computational Mathematics Science and Engineering (CMSE). She received her B.S. degrees in pharmaceutical science and psychology from Peking University, China, after which she moved to the US to pursue an M.S. degree in Criminology at the University of Pennsylvania. This was when she became fascinated by the commonality between social and biological systems and the parallel problems of control in criminal justice and medicine, for which she pursued a Ph.D. in Complex Systems and Brain Sciences at Florida Atlantic University. During postdoc positions at Stanford University and University at North Carolina – Chapel Hill, her research direction became focused on developing modeling and data analysis tools better tailored for high-dimensional dynamical systems such as neural and social systems, which are still lacking in the field. Now, her lab lives in the intersection between dynamical systems, computational topology, and neuroscience.