tag:blogger.com,1999:blog-10448119.post3437160815310578442..comments2024-03-23T06:42:53.608-04:00Comments on Hogg's Research: objective filter design, BayesHogghttp://www.blogger.com/profile/18398397408280534592noreply@blogger.comBlogger3125tag:blogger.com,1999:blog-10448119.post-42906389669115344122013-11-13T17:56:41.693-05:002013-11-13T17:56:41.693-05:00To be fair, I might have been misusing it too...or...To be fair, I might have been misusing it too...or there could be several variants of similar ideas with that name.Iain Murrayhttps://www.blogger.com/profile/02208307937369256100noreply@blogger.comtag:blogger.com,1999:blog-10448119.post-42106889986722973142013-11-13T17:44:27.806-05:002013-11-13T17:44:27.806-05:00I might be mis-using the term "multi-canonica...I might be mis-using the term "multi-canonical". Will find out.Hogghttps://www.blogger.com/profile/18398397408280534592noreply@blogger.comtag:blogger.com,1999:blog-10448119.post-34305706006650485992013-11-13T17:09:55.079-05:002013-11-13T17:09:55.079-05:00I'd be interested in any new perspectives from...I'd be interested in any new perspectives from a careful comparison in a real scenario.<br /><br />There's a toy comparison in my PhD thesis. From memory multicanonical is theoretically amazing if you can sample from it, and know the weights. But learning the weights seemed hard. And sampling from its distribution seemed hard.Iain Murrayhttps://www.blogger.com/profile/02208307937369256100noreply@blogger.com