Comments on: Asymptotically Exact, Embarrassingly Parallel MCMC
https://xianblog.wordpress.com/2013/11/26/asymptotically-exact-embarrassingly-parallel-mcmc/
an attempt at bloggin, nothing more...Sun, 16 Mar 2014 10:51:13 +0000
hourly
1 http://wordpress.com/
By: Running into a Stan Reference by Accident « Statistical Modeling, Causal Inference, and Social Science Statistical Modeling, Causal Inference, and Social Science
https://xianblog.wordpress.com/2013/11/26/asymptotically-exact-embarrassingly-parallel-mcmc/comment-page-1/#comment-46568
Sun, 16 Mar 2014 10:51:13 +0000http://xianblog.wordpress.com/?p=22433#comment-46568[…] a neat paper, which Xi’an already blogged about months ago. But what really struck me was the following […]
]]>
By: Advances in scalable Bayesian computation [day #4] ← Patient 2 Earn
https://xianblog.wordpress.com/2013/11/26/asymptotically-exact-embarrassingly-parallel-mcmc/comment-page-1/#comment-45590
Sat, 08 Mar 2014 06:43:45 +0000http://xianblog.wordpress.com/?p=22433#comment-45590[…] blog.) And in the afternoon session, Sylvia Frühwirth-Schnatter exposed her approach to the (embarrassingly) parallel problem, in the spirit of Steve’s , David Dunson’s and Scott’s (a paper posted on […]
]]>
By: Advances in scalable Bayesian computation [day #3] ← Patient 2 Earn
https://xianblog.wordpress.com/2013/11/26/asymptotically-exact-embarrassingly-parallel-mcmc/comment-page-1/#comment-45467
Fri, 07 Mar 2014 06:48:43 +0000http://xianblog.wordpress.com/?p=22433#comment-45467[…] and images. (Check the video at 45:00.) Then Steve Scott gave us a Google motivated entry to embarrassingly parallel algorithms, along the lines of papers recently discussed on the ‘Og. (Too bad we forgot to start the […]
]]>
By: xi'an
https://xianblog.wordpress.com/2013/11/26/asymptotically-exact-embarrassingly-parallel-mcmc/comment-page-1/#comment-40384
Tue, 26 Nov 2013 06:06:41 +0000http://xianblog.wordpress.com/?p=22433#comment-40384True, there is nothing that prevents you from applying this idea to ABC, assuming the product decomposition leads to a proper “subposterior” distribution for each term. But you could also get closer to ABC by (a) create a reference table of prior simulations for all threads; (b) create a reference table of pseudo-sub-samples on each thread; (c) compute the respective part of the distance between sample and pseudo-sample on each thread (and throw away the reference table overboard); (d) get back together all thread distance components and select the ABC subsample.
]]>
By: Dan Simpson
https://xianblog.wordpress.com/2013/11/26/asymptotically-exact-embarrassingly-parallel-mcmc/comment-page-1/#comment-40380
Tue, 26 Nov 2013 01:19:57 +0000http://xianblog.wordpress.com/?p=22433#comment-40380I wonder if this could be combined with ABC-like ideas. Splitting the data is all well and good when it’s informative (which is, in some sense, the less interesting big data Bayes case), but when it’s not, it may be better to do a more careful decomposition step. (The question, I guess, is what sort of models would let you do this legally…)
]]>