Dynamical Heterogeneities in Polymers

Machine Learning for Predicting Dynamical Heterogeneities in Polymers


In this project, trainees will develop predictive models of single-chain polymer dynamics through a combination of molecular dynamics (MD) simulations and ML methods, with the long-term goal of elucidating and predicting dynamical heterogeneities in polymers. Students will get training in simulating polymer systems using coarse-grained (CG) models and applying classification, dimensionality-reduction, and deep learning tools to most efficiently relate local dynamics to local structure.

Flow Chart that describes research process. Step 1 is material exploration, step 2 is data generation, step 3 is ML prediction