Deep Learning for Spatio-Temporal Flows
Prediction and uncertainty quantification of spatio-temporal flows is a challenging problem as dynamic spatio-temporal data possess underlying complex interactions and nonlinearities. Traditional statistical modeling approaches use a data generating process, generally motivated by physical laws or constraints. With examples in traffic and high frequency trading, this talk explores the wider implications for scientific research when such interactions and nonlinearities can be captured without using a data generating process.
Understanding Biomolecular Mechanisms With Molecular Dynamics Simulations
The physical functions and interactions of biological macromolecules form the basis for the diverse mechanisms that occur in the cell and underpin the processes that make life possible. To better understand the physical basis for how protein, DNA, and other biological macromolecules function, our group utilizes conventional and advanced molecular dynamics simulations to provide atomic-level descriptions for the kinetics and thermodynamics of systems of interest. In this talk, I will describe how we are using these methods on a variety of systems, as well as the challenges and opportunities they present.