Dresden 2014 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
O: Fachverband Oberflächenphysik
O 85: Molecular Simulations
O 85.5: Talk
Thursday, April 3, 2014, 18:45–19:00, WIL B321
Free energy surface reconstruction from umbrella samples using Gaussian process regression — •Thomas Stecher1,2, Noam Bernstein3, and Gábor Csányi1 — 1Department Chemie, TU München, Garching, Deutschland — 2Department of Engineering, University of Cambridge, Cambridge, UK — 3Naval Research Laboratory, Center for Computational Materials Science, Washington, DC, USA
We demonstrate how a prior assumption of smoothness can be used to enhance the reconstruction of free energy profiles from multiple umbrella sampling simulations using the Bayesian Gaussian process regression approach. The method we derive allows the concurrent use of histograms and free energy gradients and can easily be extended to include further data. In a system with one collective variable we demonstrate improved performance with respect to the weighted histogram analysis method and obtain meaningful error bars without any significant additional computation. In the case of multiple collective variables we compare to a reconstruction using least squares fitting of radial basis functions and find substantial improvements in the regimes of spatially sparse data or short sampling trajectories.