Robert Q. Topper,
Chair and Professor of Chemistry
212.353.4378
robert.topper@cooper.edu
Faculty Bio
Machine learning, Monte Carlo and molecular dynamics
The prediction of thermodynamic properties at the molecular level using Monte Carlo and molecular dynamics simulations is an active research area in our group.
Our early papers with Truhlar on Feynman path-integral adaptive stratified sampling Monte Carlo calculations of molecular thermodynamic properties are summarized in a volume of Advances in Chemical Physics. See here.
We continue to support and update TransRot, our group’s machine-portable software for carrying out simulated annealing Monte Carlo optimizations of these systems. Please visit https://github.com/steventopper/Transrot .
Recent work has included a collaboration with Mark Tuckerman (New York U), focused on the application of machine learning techniques (neural networks, Gaussian kernel ridge regression...) to free energy surfaces obtained from ergodic MD simulations. That work was published in Phys Rev Letts and J Phys Chem B.