Measuring sample quality with diffusions
- π€ Speaker: Sebastian Vollmer (Oxford)
- π Date & Time: Friday 25 November 2016, 16:00 - 17:00
- π Venue: MR12, Centre for Mathematical Sciences, Wilberforce Road, Cambridge.
Abstract
Standard Markov chain Monte Carlo diagnostics, like effective sample size, are ineffective for biased sampling procedures that sacrifice asymptotic correctness for computational speed. Recent work addresses this issue for a class of strongly log-concave target distributions by constructing a computable discrepancy measure based on Steinβs method that provably determines convergence to the target. We generalize this approach to cover any target with a fast-coupling Ito diffusion by bounding the derivatives of Stein equation solutions in terms of Markov process coupling rates. As example applications, we develop computable and convergence-determining diffusion Stein discrepancies for log-concave, heavy-tailed, and multimodal targets and use these quality measures to select the hyperparameters of biased samplers, compare random and deterministic quadrature rules, and quantify bias-variance tradeoffs in approximate Markov chain Monte Carlo. Our explicit multivariate Stein factor bounds may be of independent interest. Preprint: https://arxiv.org/abs/1611.06972
Series This talk is part of the Statistics series.
Included in Lists
- All CMS events
- All Talks (aka the CURE list)
- bld31
- Cambridge Forum of Science and Humanities
- Cambridge Language Sciences
- Cambridge talks
- Chris Davis' list
- CMS Events
- custom
- DPMMS info aggregator
- DPMMS lists
- DPMMS Lists
- Guy Emerson's list
- Hanchen DaDaDash
- Interested Talks
- Machine Learning
- MR12, Centre for Mathematical Sciences, Wilberforce Road, Cambridge.
- rp587
- School of Physical Sciences
- Statistical Laboratory info aggregator
- Statistics
- Statistics Group
Note: Ex-directory lists are not shown.
![[Talks.cam]](/static/images/talkslogosmall.gif)

Sebastian Vollmer (Oxford)
Friday 25 November 2016, 16:00-17:00