Backward Simulation of Stochastic Process Using a Time Reverse Monte Carlo Method

Abstract

The "backward simulation" of a stochastic process is defined as the stochastic dynamics that trace a time-reversed path from the target region to the initial configuration. If the probabilities calculated by the original simulation are easily restored from those obtained by backward dynamics, we can use it as a computational tool. It is shown that the naïve approach to backward simulation does not work as expected. As a remedy, the time reverse Monte Carlo method (TRMC) based on the ideas of sequential importance sampling (SIS) and sequential Monte Carlo (SMC) is proposed and successfully tested with a stochastic typhoon model and the Lorenz 96 model. TRMC with SMC, which contains resampling steps, is found to be more efficient for simulations with a larger number of time steps. A limitation of TRMC and its relation to the Bayes formula are also discussed. * stakaya@ism.ac.jp

Simulating Diffusion Bridges with Score Matching

“…Extending this work to facilitate statistical inference in the regime of partial and noisy observations could be considered along the lines of Golightly and Wilkinson (2008). Lastly, we note that our time-reversal methodology can be used to design the backward dynamics required by Takayanagi and Iba (2018) to efficiently estimate the probability of rare events driven by diffusion processes.…”

Exploring Configuration Space and Path Space of Biomolecules Using Enhanced Sampling Techniques—Searching for Mechanism and Kinetics of Biomolecular Functions

“…As mentioned above, MSES can be combined with the OM action to sample path space of a model polymer [35]. Some researchers used the OM action and the other actions for reweighting in path space [80,[82][83][84]. By minimizing or optimizing the OM action and the other actions, we can obtain a "most probable" path, and such a strategy was used in [85,86].…”

Exploring Configuration Space and Path Space of Biomolecules Using Enhanced Sampling Techniques—Searching for Mechanism and Kinetics of Biomolecular Functions

“…As mentioned above, MSES can be combined with the OM action to sample path space of a model polymer [ 37 ]. Some researchers used the OM action and the other actions for reweighting in path space [ 93 , 94 , 95 , 96 , 97 ]. By minimizing or optimizing the OM action and the other actions, we can obtain a “most probable” path, and such a strategy was used in [ 98 , 99 ].…”