ABSTRACT:
Computing the energy of a molecular system and the forces acting on the atoms is a very expensive computational task, which scales very badly with molecular size. It requires solving the Schrödinger equation (or some related cousins). We will present a Neural Network model that has learned to mimic the results of a true quantum chemical calculation, but in orders of magnitude faster. We will discuss the details of the network, its training, and the results.
Bio:
Professor Adrian Roitberg was born and raised in Buenos Aires, Argentina, where he obtained his BS from the University of Buenos Aires. He earned his PhD at the Univ of Illinois at Chicago, followed by a postdoc at Northwestern University and a stint at NIST. He joined UF in 2001, where he is now the V.T. and Louise Jackson Professor of Chemistry.