Metropolis-Coupled MCMC for Nested Sampling

Add to your list(s) Download to your calendar using vCal

• Ratthachat (Jung)
• Wednesday 07 March 2007, 14:00-15:00
• TCM Seminar Room, Cavendish Laboratory, Department of Physics.

If you have a question about this talk, please contact David MacKay.

In this talk, I will first introduce nested sampling, which is a recent probabilistic numerical integration algorithm.

After that, I’ll focus on the problem of sampling a point from a truncated domain which is essential to the accuracy of nested sampling. To solve this problem, I propose the technique called Metropolis-coupled MCMC (MCMCMC) which can unify a random-walk MCMC and the ellipsoid method.

The key idea is to use the ellipsoid method to create an auxiliary distribution. The auxiliary distribution then help the random-walk MCMC to converge faster. The MCMCMC framework is general and theoretically correct as the MCMC framework.

This talk is part of the Inference Group series.

This talk is included in these lists:

• All Cavendish Laboratory Seminars
• All Talks (aka the CURE list)
• Biology
• Cambridge Neuroscience Seminars
• Cambridge talks
• Centre for Health Leadership and Enterprise
• Chris Davis' list
• Featured lists
• Guy Emerson's list
• Hanchen DaDaDash
• Inference Group
• Inference Group Summary
• Interested Talks
• Joint Machine Learning Seminars
• Life Science
• Life Sciences
• ME Seminar
• ML
• Machine Learning Summary
• Neurons, Fake News, DNA and your iPhone: The Mathematics of Information
• Neuroscience
• Neuroscience Seminars
• Neuroscience Seminars
• Required lists for MLG
• School of Physical Sciences
• Stem Cells & Regenerative Medicine
• TCM Seminar Room, Cavendish Laboratory, Department of Physics
• Thin Film Magnetic Talks
• dh539
• kt532's list
• rp587
• yk373's list

Note that ex-directory lists are not shown.