Fast inference with min-sum matrix product

Date

Authors

Felzenszwalb, Pedro F.

McAuley, Julian

Publisher

Institute of Electrical and Electronics Engineers (IEEE Inc)

Abstract

The MAP inference problem in many graphical models can be solved efficiently using a fast algorithm for computing min-sum products of n × n matrices. The class of models in question includes cyclic and skip-chain models that arise in many applications. A

Keywords

Keywords: Expected time; Fast algorithms; Fast inference; GraphicaL model; Inference problem; Input matrices; Matrix products; Min-sum; NAtural language processing; Performance Gain; Product operations; Uniform distribution; Worst-case complexity; Algorithms; Compu Graphical models; MAP inference; min-sum matrix product

URI

http://hdl.handle.net/1885/52265

Collections

ANU Research Publications

Source

IEEE Transactions on Pattern Analysis and Machine Intelligence

Type

Journal article

DOI

10.1109/TPAMI.2011.121

Restricted until

2037-12-31

Downloads

File

Description

01_Felzenszwalb_Fast_inference_with_min-sum_2011.pdf (667.32 KB)