Monte Carlo estimation of diffusion distributions at inter-sampling times

MaRDI QIDQ4227224zbMATH OpenFDO

Authors Christian J. Wypasek, Jeffrey V. Butera, Ben G. Fitzpatrick

Publication date 12 July 1999

zbMATH Keywords

back-propagationdiffusion distributionsestimationMarkov chainMonte Carlo simulation

Mathematics Subject Classification ID

Diffusion processes (60J60) Monte Carlo methods (65C05) Stochastic systems and control (93E99)

Review text

This paper presents a back-propagation technique for estimating diffusion distributions given an initial distribution and a desired terminal distribution. The algorithm presented involves a pathwise analysis for a general diffusion. In the simplest case of a random walk on a grid, one can enumerate paths and back-propagate path probabilities in a way that is reminiscent of neural networks. The algorithm's run time and storage requirements can quickly exceed practical limits. For more general diffusions, the algorithm can be executed via Monte Carlo simulation with minimal storage requirements. For the entire collection see [Zbl 0905.00042].

Recommendations

• Importance sampling techniques for estimation of diffusion models
• Monte Carlo simulation of nonlinear diffusion processes
• Exact simulation of diffusions
• Efficient Estimation of One-Dimensional Diffusion First Passage Time Densities via Monte Carlo Simulation
• Retrospective exact simulation of diffusion sample paths with applications

Cited in

(3)

• Efficient Estimation of One-Dimensional Diffusion First Passage Time Densities via Monte Carlo Simulation
• scientific article; zbMATH DE number 2050851 (Why is no real title available?)
• Monte-Carlo estimation of time-dependent statistical characteristics of random dynamical systems

This page was built for publication: Monte Carlo estimation of diffusion distributions at inter-sampling times

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4227224)