REVIEW 5 cited by
No Free Lunch for Approximate MCMC
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
No Free Lunch for Approximate MCMC
read the original abstract
It is widely known that the performance of Markov chain Monte Carlo (MCMC) can degrade quickly when targeting computationally expensive posterior distributions, such as when the sample size is large. This has motivated the search for MCMC variants that scale well to large datasets. One popular general approach has been to look at only a subsample of the data at every step. In this note, we point out that well-known MCMC convergence results often imply that these ``subsampling'' MCMC algorithms cannot greatly improve performance. We apply these abstract results to realistic statistical problems and proposed algorithms, and also discuss some design principles suggested by the results. Finally, we develop estimates for the singular values of random matrices bounds that may be of independent interest.
Forward citations
Cited by 5 Pith papers
-
Variational predictive resampling
Variational predictive resampling uses sequential imputation from variational predictives to generate samples whose distribution converges to the exact Bayesian posterior in Gaussian models and improves dependence cap...
-
Variational predictive resampling
Variational predictive resampling iteratively imputes data from a variational predictive to produce posterior samples that converge to the exact Bayesian posterior in Gaussian models where mean-field VI retains a gap.
-
Accurate and Efficient MCMC for Latent Position Models
Two MCMC algorithms for latent position models with almost O(|E|) and O(|V|) running times plus stronger accuracy guarantees than Rastelli et al. (2024).
-
Stabilised weighted data subsampling for accelerated inference in models with recursive likelihoods
Stabilised weighted subsampling yields an unbiased log-likelihood estimator for recursive models that reduces recursion depth and computational cost while avoiding variance inflation via principled decay restrictions.
-
Stabilised weighted data subsampling for accelerated inference in models with recursive likelihoods
Stabilised weighted subsampling yields unbiased log-likelihood and gradient estimators for faster inference in recursive likelihood models with controlled variance via hyperparameter tuning.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.