Adaptive methods for efficient sampling in molecular dynamic simulations
Estadístiques de LA Referencia / Recolecta
Inclou dades d'ús des de 2022
Cita com:
hdl:2117/333989
Tipus de documentText en actes de congrés
Data publicació2015
EditorCIMNE
Condicions d'accésAccés obert
Tots els drets reservats. Aquesta obra està protegida pels drets de propietat intel·lectual i
industrial corresponents. Sense perjudici de les exempcions legals existents, queda prohibida la seva
reproducció, distribució, comunicació pública o transformació sense l'autorització del titular dels drets
Abstract
I will present some numerical challenges raised by the simulation of materials at the atomistic level. One of the main difficulties is related to a timescale problem: at the molecular level, the typical timescale is much smaller than the timescale of interest at the macroscopic level. This is related to the fact that the stochastic process which models the evolution of the molecular system is metastable: it remains trapped for very long times in some regions of the phase space (called metastable states). Specific techniques have been devised in order to circumvent these problems, in order to sample efficiently the configurations and the trajectories. In this presentation, I will focus on the configurational sampling problem. The question is then how to devise efficient sampling technique to sample multimodal measures in high dimension. I will present in particular adaptive biasing techniques which have been proposed in the molecular dynamics community to construct on the fly biasing potential, in order to more efficiently sample the target measure. These techniques can thus be seen as adaptive importance sampling methods.
CitacióLelièvre, T. Adaptive methods for efficient sampling in molecular dynamic simulations. A: ADMOS 2015. CIMNE, 2015, p. 74.
Fitxers | Descripció | Mida | Format | Visualitza |
---|---|---|---|---|
Admos2015-53-Ad ... for efficient sampling.pdf | 156,2Kb | Visualitza/Obre |