On the role of interaction in sequential Monte Carlo algorithms
From MaRDI portal
Signal detection and filtering (aspects of stochastic processes) (60G35) Interacting random processes; statistical mechanics type models; percolation theory (60K35) Markov processes: estimation; hidden Markov models (62M05) Monte Carlo methods (65C05) Numerical analysis or methods applied to Markov chains (65C40)
Abstract: We introduce a general form of sequential Monte Carlo algorithm defined in terms of a parameterized resampling mechanism. We find that a suitably generalized notion of the Effective Sample Size (ESS), widely used to monitor algorithm degeneracy, appears naturally in a study of its convergence properties. We are then able to phrase sufficient conditions for time-uniform convergence in terms of algorithmic control of the ESS, in turn achievable by adaptively modulating the interaction between particles. This leads us to suggest novel algorithms which are, in senses to be made precise, provably stable and yet designed to avoid the degree of interaction which hinders parallelization of standard algorithms. As a byproduct, we prove time-uniform convergence of the popular adaptive resampling particle filter.
Recommendations
- On adaptive resampling strategies for sequential Monte Carlo methods
- Simple conditions for convergence of sequential Monte Carlo genealogies with applications
- On the convergence of adaptive sequential Monte Carlo methods
- Limit theorems for weighted samples with applications to sequential Monte Carlo methods
- Central limit theorem for sequential Monte Carlo methods and its application to Bayesian inference
Cites work
- A nonasymptotic theorem for unnormalized Feynman-Kac particle models
- Blind Deconvolution via Sequential Imputations
- Decentralized Particle Filter With Arbitrary State Decomposition
- Filtering via Simulation: Auxiliary Particle Filters
- scientific article; zbMATH DE number 2106098 (Why is no real title available?)
- Limit theorems for weighted samples with applications to sequential Monte Carlo methods
- Mean field simulation for Monte Carlo integration
- On adaptive resampling strategies for sequential Monte Carlo methods
- On parallel implementation of sequential Monte Carlo methods: the island particle model
- Particle filters
- Particle filters with random resampling times
- Particle Markov chain Monte Carlo for efficient numerical simulation
- Resampling algorithms and architectures for distributed particle filters
- Sequential Monte Carlo Samplers
- Stability and uniform approximation of nonlinear filters using the Hilbert metric and application to particle filters
- Stability properties of some particle filters
- Twisted particle filters
Cited in
(17)- The sample size required in importance sampling
- Sequential Monte Carlo as approximate sampling: bounds, adaptive resampling via \(\infty\)-ESS, and an application to particle Gibbs
- Importance sampling: intrinsic dimension and computational cost
- On resampling schemes for particle filters with weakly informative observations
- Sequential ensemble transform for Bayesian inverse problems
- Asymptotic genealogies of interacting particle systems with an application to sequential Monte Carlo
- A fast particle-based approach for calibrating a 3-D model of the Antarctic ice sheet
- Negative association, ordering and convergence of resampling methods
- Fluctuations, stability and instability of a distributed particle filter with local exchange
- Sequential Bayesian inference for implicit hidden Markov models and current limitations
- A sequential particle algorithm that keeps the particle system alive
- Forest resampling for distributed sequential Monte Carlo
- Smoothing with couplings of conditional particle filters
- An Invitation to Sequential Monte Carlo Samplers
- On parallel implementation of sequential Monte Carlo methods: the island particle model
- Rethinking the Effective Sample Size
- Adaptive tuning of Hamiltonian Monte Carlo within sequential Monte Carlo
This page was built for publication: On the role of interaction in sequential Monte Carlo algorithms
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5963509)
