Efficient strategy for the Markov chain Monte Carlo in high-dimension with heavy-tailed target probability distribution
From MaRDI portal
Publication:1750100
DOI10.3150/17-BEJ976zbMATH Open1407.60102arXiv1412.6231MaRDI QIDQ1750100FDOQ1750100
Authors: Kengo Kamatani 
Publication date: 18 May 2018
Published in: Bernoulli (Search for Journal in Brave)
Abstract: The purpose of this paper is to introduce a new Markov chain Monte Carlo method and exhibit its efficiency by simulation and high-dimensional asymptotic theory. Key fact is that our algorithm has a reversible proposal transition kernel, which is designed to have a heavy-tailed invariant probability distribution. The high-dimensional asymptotic theory is studied for a class of heavy-tailed target probability distribution. As the number of dimension of the state space goes to infinity, we will show that our algorithm has a much better convergence rate than that of the preconditioned Crank Nicolson (pCN) algorithm and the random-walk Metropolis (RWM) algorithm. We also show that our algorithm is at least as good as the pCN algorithm and better than the RWM algorithm for light-tailed target probability distribution.
Full work available at URL: https://arxiv.org/abs/1412.6231
Recommendations
- scientific article; zbMATH DE number 6781368
- Markov Chain Importance Sampling—A Highly Efficient Estimator for MCMC
- An efficient Markov chain Monte Carlo method for distributions with intractable normalising constants
- scientific article; zbMATH DE number 5520724
- Unbiased Markov chain Monte Carlo for intractable target distributions
- Efficient Bernoulli factory Markov chain Monte Carlo for intractable posteriors
- A new strategy for speeding Markov chain Monte Carlo algorithms
- Efficient implementation of Markov chain Monte Carlo when using an unbiased likelihood estimator
Convergence of probability measures (60B10) Computational methods in Markov chains (60J22) Monte Carlo methods (65C05) Stochastic calculus of variations and the Malliavin calculus (60H07)
Cites Work
- Weak convergence and optimal scaling of random walk Metropolis algorithms
- Markov chains for exploring posterior distributions. (With discussion)
- The Malliavin Calculus and Related Topics
- Title not available (Why is that?)
- Title not available (Why is that?)
- Multivariate T-Distributions and Their Applications
- Optimal Scaling of Discrete Approximations to Langevin Diffusions
- Normal Approximation by Stein’s Method
- Title not available (Why is that?)
- Title not available (Why is that?)
- Stein's method on Wiener chaos
- Normal approximations with Malliavin calculus. From Stein's method to universality
- Optimal scalings for local Metropolis-Hastings chains on nonproduct targets in high dimensions
- Spectral gaps for a Metropolis-Hastings algorithm in infinite dimensions
- MCMC methods for functions: modifying old algorithms to make them faster
- Derivatives of Wiener functionals and absolute continuity of induced measures
- Hybrid multi-step estimators for stochastic differential equations based on sampled data
- Title not available (Why is that?)
- Ergodicity of Markov chain Monte Carlo with reversible proposal
- Error bounds for Metropolis-Hastings algorithms applied to perturbations of Gaussian measures in high dimensions
- Local consistency of Markov chain Monte Carlo methods
- Hybrid multi-step estimation of the volatility for stochastic regression models
Cited In (18)
- Hybrid estimation for ergodic diffusion processes based on noisy discrete observations
- Hybrid estimators for stochastic differential equations from reduced data
- How to combine fast heuristic Markov chain Monte Carlo with slow exact sampling
- Hybrid estimators for small diffusion processes based on reduced data
- Estimation of risk contributions with MCMC
- Title not available (Why is that?)
- Title not available (Why is that?)
- A stable manifold MCMC method for high dimensions
- Ergodicity of Markov chain Monte Carlo with reversible proposal
- Potential-decomposition strategy in Markov chain Monte Carlo sampling algorithms
- Monte Carlo on manifolds in high dimensions
- Adaptation of the tuning parameter in general Bayesian inference with robust divergence
- Quasi-likelihood analysis and Bayes-type estimators of an ergodic diffusion plus noise
- Random walk Metropolis algorithm in high dimension with non-Gaussian target distributions
- Sequential Kalman tuning of the \(t\)-preconditioned Crank-Nicolson algorithm: efficient, adaptive and gradient-free inference for Bayesian inverse problems
- Non-reversible guided Metropolis kernel
- Title not available (Why is that?)
- High-dimensional scaling limits of piecewise deterministic sampling algorithms
Uses Software
This page was built for publication: Efficient strategy for the Markov chain Monte Carlo in high-dimension with heavy-tailed target probability distribution
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1750100)
