Efficient SMC^2 schemes for stochastic kinetic models
From MaRDI portal
Publication:2329744
Abstract: Fitting stochastic kinetic models represented by Markov jump processes within the Bayesian paradigm is complicated by the intractability of the observed data likelihood. There has therefore been considerable attention given to the design of pseudo-marginal Markov chain Monte Carlo algorithms for such models. However, these methods are typically computationally intensive, often require careful tuning and must be restarted from scratch upon receipt of new observations. Sequential Monte Carlo (SMC) methods on the other hand aim to efficiently reuse posterior samples at each time point. Despite their appeal, applying SMC schemes in scenarios with both dynamic states and static parameters is made difficult by the problem of particle degeneracy. A principled approach for overcoming this problem is to move each parameter particle through a Metropolis-Hastings kernel that leaves the target invariant. This rejuvenation step is key to a recently proposed SMC algorithm, which can be seen as the pseudo-marginal analogue of an idealised scheme known as iterated batch importance sampling. Computing the parameter weights in SMC requires running a particle filter over dynamic states to unbiasedly estimate the intractable observed data likelihood contributions at each time point. In this paper, we propose to use an auxiliary particle filter inside the SMC scheme. Our method uses two recently proposed constructs for sampling conditioned jump processes and we find that the resulting inference schemes typically require fewer state particles than when using a simple bootstrap filter. Using two applications, we compare the performance of the proposed approach with various competing methods, including two global MCMC schemes.
Recommendations
- Efficient stochastic model for the point kinetics equations
- Efficient kinetic Monte Carlo method for reaction-diffusion problems with spatially varying annihilation rates
- Efficient stochastic Galerkin methods for random diffusion equations
- Synchronous parallel kinetic Monte Carlo for continuum diffusion-reaction systems
- Correlated pseudo-marginal schemes for time-discretised stochastic kinetic models
- An efficient stochastic algorithm for solving the Boltzmann equation
- Iterative semi-implicit splitting methods for stochastic chemical kinetics
- Efficient kinetic Monte Carlo simulation
- Stochastic simulation of fluctuation-induced reaction-diffusion kinetics governed by Smoluchowski equations
- On efficient simulations of multiscale kinetic transport
Cites work
- \(\mathrm{SMC}^2\): an efficient algorithm for sequential analysis of state space models
- A hierarchy of approximations of the master equation scaled by a size parameter
- A sequential particle filter method for static models
- Alive SMC\(^{2}\): Bayesian model selection for low-count time series models with intractable likelihoods
- Bayesian inference for Markov jump processes with informative observations
- Bayesian inference for stochastic epidemic models with time-inhomogeneous removal rates
- Bayesian Inference for Stochastic Kinetic Models Using a Diffusion Approximation
- Efficient learning via simulation: a marginalized resample-move approach
- Exact Filtering for Partially Observed Continuous Time Models
- Filtering via Simulation: Auxiliary Particle Filters
- Following a moving target -- Monte Carlo inference for dynamic Bayesian models
- scientific article; zbMATH DE number 3523369 (Why is no real title available?)
- scientific article; zbMATH DE number 1180828 (Why is no real title available?)
- scientific article; zbMATH DE number 2106098 (Why is no real title available?)
- On some properties of Markov chain Monte Carlo simulation methods based on the particle filter
- On the efficiency of pseudo-marginal random walk Metropolis algorithms
- Particle Gibbs with ancestor sampling
- Particle learning and smoothing
- Particle Markov Chain Monte Carlo Methods
- Sequential Bayesian inference for implicit hidden Markov models and current limitations
- Sequential Bayesian inference in hidden Markov stochastic kinetic models with application to detection and response to seasonal epidemics
- Sequential Monte Carlo Methods in Practice
- Simulation-based Bayesian inference for epidemic models
- Stochastic epidemic models and their statistical analysis
- Stochastic modelling for systems biology.
- The alive particle filter and its use in particle Markov chain Monte Carlo
- The correlated pseudomarginal method
- Twisted particle filters
Cited in
(12)- Bayesian model discrimination for partially-observed epidemic models
- Importance sampling for partially observed temporal epidemic models
- Efficient sampling of conditioned Markov jump processes
- Correlated pseudo-marginal schemes for time-discretised stochastic kinetic models
- Bayesian computation methods for inference in stochastic kinetic models
- Delayed acceptance particle MCMC for exact inference in stochastic kinetic models
- \(\mathrm{SMC}^2\): an efficient algorithm for sequential analysis of state space models
- A population Monte Carlo scheme with transformed weights and its application to stochastic kinetic models
- Automatically adapting the number of state particles in \(\text{SMC}^2\)
- Exact Bayesian inference for fitting stochastic epidemic models to partially observed incidence data
- Nested ensemble Kalman filter for static parameter inference in nonlinear state-space models
- Bayesian parameter inference for partially observed stopped processes
This page was built for publication: Efficient \(\mathrm{SMC}^2\) schemes for stochastic kinetic models
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2329744)
