Fluctuations, stability and instability of a distributed particle filter with local exchange
arXiv:1505.02390
Abstract
We study a distributed particle filter proposed by Bolić et al.~(2005). This algorithm involves groups of particles, with interaction between groups occurring through a "local exchange" mechanism. We establish a central limit theorem in the regime where is fixed and . A formula we obtain for the asymptotic variance can be interpreted in terms of colliding Markov chains, enabling analytic and numerical evaluations of how the asymptotic variance behaves over time, with comparison to a benchmark algorithm consisting of independent particle filters. We prove that subject to regularity conditions, when is fixed both algorithms converge time-uniformly at rate . Through use of our asymptotic variance formula we give counter-examples satisfying the same regularity conditions to show that when is fixed neither algorithm, in general, converges time-uniformly at rate .
arXiv comment49 pages, 7 figures
References in corpus (5)
- Central limit theorem for sequential Monte Carlo methods and its application to Bayesian inference
- On adaptive resampling strategies for sequential Monte Carlo methods
- On the role of interaction in sequential Monte Carlo algorithms
- A proof of uniform convergence over time for a distributed particle filter
- Butterfly resampling: asymptotics for particle filters with constrained interactions