Antithetic Multilevel Particle Filters

01/29/2023
by   Ajay Jasra, et al.
0

In this paper we consider the filtering of partially observed multi-dimensional diffusion processes that are observed regularly at discrete times. This is a challenging problem which requires the use of advanced numerical schemes based upon time-discretization of the diffusion process and then the application of particle filters. Perhaps the state-of-the-art method for moderate dimensional problems is the multilevel particle filter of <cit.>. This is a method that combines multilevel Monte Carlo and particle filters. The approach in that article is based intrinsically upon an Euler discretization method. We develop a new particle filter based upon the antithetic truncated Milstein scheme of <cit.>. We show that for a class of diffusion problems, for ϵ>0 given, that the cost to produce a mean square error (MSE) in estimation of the filter, of 𝒪(ϵ^2) is 𝒪(ϵ^-2log(ϵ)^2). In the case of multidimensional diffusions with non-constant diffusion coefficient, the method of <cit.> has a cost of 𝒪(ϵ^-2.5) to achieve the same MSE. We support our theory with numerical results in several examples.

READ FULL TEXT

Please sign up or login with your details

Forgot password? Click here to reset