Faster Exact MPE and Constrained Optimization with Deterministic Finite State Automata

TL;DR

This paper introduces FABE, a novel approach using deterministic finite automata to enhance Bucket Elimination for exact MPE and constrained optimization, achieving significant speed and memory improvements.

Contribution

The paper presents a new concise automata-based representation integrated into Bucket Elimination, substantially improving efficiency for graphical model inference tasks.

Findings

FABE outperforms existing methods in runtime by up to 5 orders of magnitude.

The approach reduces memory usage compared to traditional Bucket Elimination.

Experimental results demonstrate effectiveness on benchmark problems.

Abstract

We propose a concise function representation based on deterministic finite state automata for exact most probable explanation and constrained optimization tasks in graphical models. We then exploit our concise representation within Bucket Elimination (BE). We denote our version of BE as FABE. FABE significantly improves the performance of BE in terms of runtime and memory requirements by minimizing redundancy. Results on most probable explanation and weighted constraint satisfaction benchmarks show that FABE often outperforms the state of the art, leading to significant runtime improvements (up to 5 orders of magnitude in our tests).