Committee Members:

Prof. Rampi Ramprasad, Advisor, MSE

Prof. Seung Soon Jang, Co-Advisor, MSE

Prof. Karl I. Jacob, MSE

Prof. Ryan P. Lively, ChBE

Prof. Chao Zhang, CSE

Methodological Developments for Polymer Informatics

Abstract:

Finding a polymeric material tailored to a specific application constitutes a daunting search problem, given the staggeringly large polymer chemical space. In these scenarios, the use of machine learning (ML) models to rapidly screen polymers and design for desired performances has become a powerful approach.

In this work, we use ML to study and design polymers for gas separation membranes and for dielectrics. Good ML models require sufficient data to train. As such, one aspect of this work is data collection. Another aspect is the development of new methods that advance the speed and accuracy of polymer informatics. These developments will touch on the numerical representation of polymers, digital synthesis planning of polymers, and property prediction. Further, we will develop ML methods that go beyond property prediction of polymers and, instead, predict polymers directly from user-desired target criteria. In other words, these methods will solve the inverse problem.