• Corpus ID: 118018

Ergodicity of Approximate MCMC Chains with Applications to Large Data Sets

@article{Pillai2014ErgodicityOA,
  title={Ergodicity of Approximate MCMC Chains with Applications to Large Data Sets},
  author={Natesh S. Pillai and Aaron Smith},
  journal={arXiv: Statistics Theory},
  year={2014}
}
In many modern applications, difficulty in evaluating the posterior density makes performing even a single MCMC step slow. This difficulty can be caused by intractable likelihood functions, but also appears for routine problems with large data sets. Many researchers have responded by running approximate versions of MCMC algorithms. In this note, we develop quantitative bounds for showing the ergodicity of these approximate samplers. We then use these bounds to study the bias-variance trade-off… 
View PDF on arXiv
58 Citations
Highly Influential Citations
4
Background Citations
31
Methods Citations
9
Results Citations
5

58 Citations

Stability and examples of some approximate MCMC algorithms.

This work examines the stability properties of a particular approximate MCMC algorithm, which is term noisy Metropolis-Hastings, and presents a novel approximate method, motivated by the class of exact algorithms with randomised acceptance.

Bounding Wasserstein distance with couplings

Quality measures are applied to stochastic gradient MCMC, variational Bayes, and Laplace approximations for tall data and to approximate MCMC for Bayesian logistic regression in 4500 dimensions and Bayesian linear regression in 50000 dimensions.

A fast asynchronous MCMC sampler for sparse Bayesian inference

It is shown that in high-dimensional linear regression problems, the Markov chain generated by the proposed algorithm admits an invariant distribution that recovers correctly the main signal with high probability under some statistical assumptions.

Approximations of Markov Chains and High-Dimensional Bayesian Inference

A framework for assessing when to use approximations in MCMC algorithms, and how much error in the transition kernel should be tolerated to obtain optimal estimation performance with respect to a specified loss function and computational budget is proposed.

Scalable Approximate MCMC Algorithms for the Horseshoe Prior

The empirical results show that the new algorithm yields estimates with lower mean squared error, intervals with better coverage, and elucidates features of the posterior that were often missed by previous algorithms in high dimensions, including bimodality of posterior marginals indicating uncertainty about which covariates belong in the model.

Optimal approximating Markov chains for Bayesian inference

This work gives simple, sharp bounds for uniform approximations of uniformly mixing Markov chains, and suggests a notion of optimality that incorporates computation time and approximation error, and uses these bounds to make generalizations about properties of good approximation in the uniformly mixing setting.

Rate-optimal refinement strategies for local approximation MCMC

It is proved that the error of the resulting local approximation MCMC (LA-MCMC) algorithm decays at roughly the expected $1/\sqrt{T}$ rate, and an algorithmic parameter is introduced that guarantees convergence given very weak tail bounds, significantly strengthening previous convergence results.

Posterior Computation with the Gibbs Zig-Zag Sampler

GZZ provides a flexible framework to combine PDMPs with the rich literature on MCMC algorithms and proves appealing theoretical properties of GZZ and demonstrates it on posterior sampling for logistic models with shrinkage priors for high-dimensional regression and random effects.

Approximations of Markov Chains and Bayesian Inference

A framework for assessing when to use approximations in MCMC algorithms, and how much error in the transition kernel should be tolerated to obtain optimal estimation performance with respect to a specified discrepancy measure and computational budget is proposed.

Light and Widely Applicable MCMC: Approximate Bayesian Inference for Large Datasets

A Markov chain whose transition makes use of an unknown but fixed, fraction of the available data, where the random choice of sub-sample is guided by the fidelity of this sub- sample to the observed data, as measured by summary (or sucient) statistics.
...

References

SHOWING 1-10 OF 49 REFERENCES

Approximations of Markov Chains and High-Dimensional Bayesian Inference

A framework for assessing when to use approximations in MCMC algorithms, and how much error in the transition kernel should be tolerated to obtain optimal estimation performance with respect to a specified loss function and computational budget is proposed.

Optimal approximating Markov chains for Bayesian inference

This work gives simple, sharp bounds for uniform approximations of uniformly mixing Markov chains, and suggests a notion of optimality that incorporates computation time and approximation error, and uses these bounds to make generalizations about properties of good approximation in the uniformly mixing setting.

Accelerating Asymptotically Exact MCMC for Computationally Intensive Models via Local Approximations

The ergodicity of the approximate Markov chain is proved, showing that it samples asymptotically from the exact posterior distribution of interest, and variations of the algorithm that employ either local polynomial approximations or local Gaussian process regressors are described.

Light and Widely Applicable MCMC: Approximate Bayesian Inference for Large Datasets

A Markov chain whose transition makes use of an unknown but fixed, fraction of the available data, where the random choice of sub-sample is guided by the fidelity of this sub- sample to the observed data, as measured by summary (or sucient) statistics.

The pseudo-marginal approach for efficient Monte Carlo computations

A powerful and flexible MCMC algorithm for stochastic simulation that builds on a pseudo-marginal method, showing how algorithms which are approximations to an idealized marginal algorithm, can share the same marginal stationary distribution as the idealized method.

Subsampling-Based Approximate Monte Carlo for Discrete Distributions

The problem of sampling a discrete random variable with a high degree of dependency that is typical in large-scale Bayesian inference and graphical models is studied, and an efficient approximate solution with a subsampling approach is proposed.

Minorization Conditions and Convergence Rates for Markov Chain Monte Carlo

Abstract General methods are provided for analyzing the convergence of discrete-time, general state-space Markov chains, such as those used in stochastic simulation algorithms including the Gibbs…

Scalable Discrete Sampling as a Multi-Armed Bandit Problem

The problem of sampling a discrete random variable with a high degree of dependency that is typical in large-scale Bayesian inference and graphical models is studied, and an efficient approximate solution with a subsampling approach is proposed.

Austerity in MCMC Land: Cutting the Metropolis-Hastings Budget

This work introduces an approximate MH rule based on a sequential hypothesis test that allows us to accept or reject samples with high confidence using only a fraction of the data required for the exact MH rule.

Variance bounding and geometric ergodicity of Markov chain Monte Carlo kernels for approximate Bayesian computation

It is proved that a recently introduced Markov kernel can inherit the properties of variance bounding and geometric ergodicity from its intractable Metropolis–Hastings counterpart, under reasonably weak conditions.