Graph Partitioning for Independent Sets

TL;DR

This paper presents an evolutionary algorithm for maximum independent set problems that uses graph partitioning and local search to improve solution quality and efficiency, outperforming existing methods.

Contribution

The paper introduces a novel evolutionary algorithm leveraging graph partitioning-based combine operations and local search, advancing the state-of-the-art in independent set computation.

Findings

Outperforms existing algorithms on various instances

Uses graph partitioning for efficient combine operations

Combines local search with evolutionary strategies

Abstract

Computing maximum independent sets in graphs is an important problem in computer science. In this paper, we develop an evolutionary algorithm to tackle the problem. The core innovations of the algorithm are very natural combine operations based on graph partitioning and local search algorithms. More precisely, we employ a state-of-the-art graph partitioner to derive operations that enable us to quickly exchange whole blocks of given independent sets. To enhance newly computed offsprings we combine our operators with a local search algorithm. Our experimental evaluation indicates that we are able to outperform state-of-the-art algorithms on a variety of instances.