The Relationship Between AND/OR Search and Variable Elimination

TL;DR

This paper introduces a unified framework for comparing search and inference algorithms in graphical models, highlighting their relationships and differences through the lens of AND/OR search.

Contribution

It provides a novel comparison of Variable Elimination and AND/OR Search, integrating various algorithms within this framework to clarify their connections.

Findings

VE and AO are closely related, with AO encompassing several advanced algorithms.

The framework reveals conditions under which different algorithms are more efficient.

Insights into memory usage and algorithm performance in graphical models.

Abstract

In this paper we compare search and inference in graphical models through the new framework of AND/OR search. Specifically, we compare Variable Elimination (VE) and memoryintensive AND/OR Search (AO) and place algorithms such as graph-based backjumping and no-good and good learning, as well as Recursive Conditioning [7] and Value Elimination [2] within the AND/OR search framework.