A variational inference procedure allowing internal structure for overlapping clusters and deterministic constraints

Donald R Geiger,Christopher Meek,Ydo Wexler, Alla A Epshteyn,Gerald Dejong,Joyce Y Chai,Zahar Prasov, Songnan Qu,Mirella Lapata,Alex Lascarides, Jungwoo Oh, K Choi,Hitoshi Isahara,Branislav Kveton,Milos Hauskrecht,Carlos Guestrin,Ion Muslea,Steven Minton,Craig A Knoblock,Maria Fox, Darrell Long, L Liu, Miroslaw Truszczynski,Joelle Pineau, Gary B Gordon,Sebastian Thrun, Shaheen Fatima,Michael Wooldridge,Nicholas R Jennings, S Hoelldobler,Eldar Karabaev,Olga Skvortsova, Yi Zhang,Roland H C Yap,Paolo Viappiani,Boi Faltings,Pearl Pu,Dmitri A Dolgov,Edmund H Durfee, Ozlem Yilmaz,A C Cem Say,Laura Barbulescu,Adele E Howe,L D Whitley, Mandy L Roberts,Francesca Rossi,Kristen Brent Venable,Neil Yorkesmith

J. Artif. Intell. Res. (JAIR)（2006）

引用 19|浏览326

暂无评分

摘要

We develop a novel algorithm, called VIP*, for structured variational approximate inference. This algorithm extends known algorithms to allow efficient multiple potential updates for overlapping clusters, and overcomes the difficulties imposed by deterministic constraints. The algorithm's convergence is proven and its applicability demonstrated for genetic linkage analysis.

查看译文

关键词

deterministic constraint,internal structure,overlapping cluster,efficient multiple potential updates,variational inference procedure,genetic linkage analysis,novel algorithm,structured variational approximate inference,genetic linkage

AI 理解论文

溯源树

样例

生成溯源树，研究论文发展脉络

Chat Paper

正在生成论文摘要