The practice and theory of stochastic simulation
October 22 to October 26, 2007
American Institute of Mathematics,
San Jose, California
Jonathan Mattingly and Eric Vanden-Eijnden
This workshop concerns approaches
for the numerical integration of stochastic systems which span
many temporal-scales. Molecular dynamics and stochastic
simulations in chemical kinetics, in particular, are two
important examples among several other which require efficient and
accurate integrators. The simulations of these systems generate many
important questions which, so far, remain mostly open. Can stochastic
differential equations (SDEs) represent a viable alternative to
thermostats such as Nose-Hoover or Andersen for the simulation of
systems in ensembles other than the microcanonical one? How to
guarantee the ergodicity of integrators for such SDEs and assess
their rate of convergence, accuracy and efficiency? How to
incorporate holonomic constraints in these contexts? How to
efficiently compute expectation and free energies using such
integrators, especially in systems displaying metastability? How to
increase the efficiency of Gillespie's stochastic simulation
algorithm for chemical kinetics systems with disparate rates?
Can stochastic differential equation be a viable alternative in
this context as well?
Activities at the workshop will focus on new developments and related issues in these areas.
Material from the workshop
A list of participants.
The workshop schedule.
A report on the workshop activities.