On Unbiased Estimation for Discretized Models
From MaRDI portal
Abstract: In this article, we consider computing expectations w.r.t. probability measures which are subject to discretization error. Examples include partially observed diffusion processes or inverse problems, where one may have to discretize time and/or space, in order to practically work with the probability of interest. Given access only to these discretizations, we consider the construction of unbiased Monte Carlo estimators of expectations w.r.t. such target probability distributions. It is shown how to obtain such estimators using a novel adaptation of randomization schemes and Markov simulation methods. Under appropriate assumptions, these estimators possess finite variance and finite expected cost. There are two important consequences of this approach: (i) unbiased inference is achieved at the canonical complexity rate, and (ii) the resulting estimators can be generated independently, thereby allowing strong scaling to arbitrarily many parallel processors. Several algorithms are presented, and applied to some examples of Bayesian inference problems, with both simulated and real observed data.
Recommendations
- Unbiased inference for discretely observed hidden Markov model diffusions
- Unbiased estimation with square root convergence for SDE models
- Unbiased multi-index Monte Carlo
- Unbiased filtering of a class of partially observed diffusions
- Unbiased Markov chain Monte Carlo for intractable target distributions
Cites work
- scientific article; zbMATH DE number 3934272 (Why is no real title available?)
- scientific article; zbMATH DE number 48727 (Why is no real title available?)
- scientific article; zbMATH DE number 1972910 (Why is no real title available?)
- scientific article; zbMATH DE number 1522714 (Why is no real title available?)
- scientific article; zbMATH DE number 1416816 (Why is no real title available?)
- A general method for debiasing a Monte Carlo estimator
- An Introduction to Numerical Analysis
- Coupling and convergence for Hamiltonian Monte Carlo
- Coupling of multidimensional diffusions by reflection
- Effective containment explains subexponential growth in recent confirmed COVID-19 cases in China
- Exact estimation for Markov chain equilibrium expectations
- Inverse Problem Theory and Methods for Model Parameter Estimation
- Inverse problems: a Bayesian perspective
- MCMC methods for functions: modifying old algorithms to make them faster
- Machine learning. A probabilistic perspective
- Markov chain simulation for multilevel Monte Carlo
- Markovian stochastic approximation with expanding projections
- Multilevel Particle Filters
- Multilevel sequential Monte Carlo samplers
- Multilevel sequential Monte Carlo with dimension-independent likelihood-informed proposals
- Smoothing with couplings of conditional particle filters
- Studying Convergence of Markov Chain Monte Carlo Algorithms Using Coupled Sample Paths
- The finite element methods for elliptic problems.
- Unbiased Hamiltonian Monte Carlo with couplings
- Unbiased Markov Chain Monte Carlo Methods with Couplings
- Unbiased Monte Carlo: posterior estimation for intractable/infinite-dimensional models
- Unbiased estimation of the gradient of the log-likelihood in inverse problems
- Unbiased estimation with square root convergence for SDE models
- Unbiased estimators and multilevel Monte Carlo
- Unbiased filtering of a class of partially observed diffusions
- Unbiased inference for discretely observed hidden Markov model diffusions
Cited in
(19)- Unbiased inference for discretely observed hidden Markov model diffusions
- Unbiased simulation of distributions with explicitly known integral transforms
- Unbiased Estimation Using Underdamped Langevin Dynamics
- Unbiased estimation using a class of diffusion processes
- Unbiased density computation for stochastic resetting
- Unbiased parameter inference for a class of partially observed Lévy-process models
- Unbiased Hamiltonian Monte Carlo with couplings
- A randomized multi-index sequential Monte Carlo method
- Asymptotic variance in the central limit theorem for multilevel Markovian stochastic approximation
- An Improved Unbiased Particle Filter
- Unbiased parameter estimation for partially observed diffusions
- Unbiased estimation of the gradient of the log-likelihood in inverse problems
- scientific article; zbMATH DE number 4106250 (Why is no real title available?)
- Unbiased parameter estimation for Bayesian inverse problems
- Optimal unbiased estimation for expected cumulative discounted cost
- Unbiased Monte Carlo: posterior estimation for intractable/infinite-dimensional models
- Unbiased approximations of products of expectations
- Unbiased approximation of posteriors via coupled particle Markov chain Monte Carlo
- Unbiased multi-index Monte Carlo
This page was built for publication: On Unbiased Estimation for Discretized Models
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6164172)
