General-Purpose Co-Evolutionary Construction of Parallel Algorithm Portfolios for Multi-Objective Binary Optimization

TL;DR

This paper introduces DACMO, a domain-agnostic co-evolutionary approach for constructing parallel algorithm portfolios tailored for multi-objective binary optimization, leveraging neural representations and LLM-based operator generation.

Contribution

It presents a novel, general-purpose method combining neural instance representation and LLM-driven operator design for multi-objective binary optimization.

Findings

DACMO outperforms classic MOEA-based portfolios across four problem classes.

DACMO matches state-of-the-art baselines without needing problem-specific instance generators.

The approach is applicable without modifications to diverse MOBOPs.

Abstract

Despite recent progress in constructing generalizable parallel algorithm portfolios (PAPs), no general-purpose approach is yet available for multi-objective binary optimization problems (MOBOPs). To fill this gap, this paper proposes domain-agnostic co-evolution of parameterized search for multi-objective binary optimization~(DACMO), which features two technical innovations. First, we propose a neural instance representation architecture that decouples domain-invariant and instance-specific features, enabling class-consistent instance generation across varying dimensions without problem-specific instance generators. Second, we introduce LLM-based automatic search operator generation into PAP construction, extending the search space from parameter tuning of predefined templates to operator-level algorithm design. We evaluate DACMO on four representative MOBOP classes to demonstrate its effectiveness as a general-purpose PAP construction method: the multi-objective match max problem~(MMMP), the multi-objective knapsack problem~(MKP), the multi-objective contamination control problem (MCCP), and the multi-objective complementary influence maximization problem~(MCIMP). Experimental results show that DACMO can be directly applied to all four problem classes without modification, outperforms PAPs built from classic MOEA templates, and achieves performance comparable to a privileged state-of-the-art baseline that relies on manually designed problem-specific instance generators, while outperforming it on two of the four evaluated problem classes.