Mark E. Tuckerman, Ph.D.,
Professor of Chemistry and Mathematics
Dept. of Chemistry and Courant Institute of Mathematical Sciences
New York University
Topic: Interfacing temperature-accelerated molecular dynamics with machine learning for generating, representing, and deploying high-dimensional free-energy landscapes
Abstract: The topic will focus on interfacing temperature-accelerated molecular dynamics for enhanced sampling with machine learning for representing high-dimensional free-energy landscapes and deploying them for the calculation of observables. A comparison of different types of machine learning models, including kernel methods, neural networks, and weighted neighbor methods will be presented. Under this topic, I also expect to discuss machine learning strategies for learning reaction coordinates from these high-dimensional landscapes.