Learning-based method decides when to use cooperative coevolution

WCC: A Learning-Based Framework for Deciding When to Apply Cooperative Coevolution in Large-Scale Optimization

Machine LearningArtificial Intelligence

Summary

Large-scale optimization problems are hard to solve because they involve many variables and different strategies work better for different problems. The authors developed WCC, a smart system that learns during optimization to decide when to use a method called Cooperative Coevolution or another approach. WCC switches between methods dynamically to save computing effort without needing experts to tell it how. It performs well across various problems and lets users train once and reuse the strategy, reducing costs and offering helpful advice on applying Cooperative Coevolution effectively.

What this means in practice

  • For machine learning engineers: Optimize large models more efficiently by dynamically choosing optimization methods during training without manual tuning.
  • For industrial optimization teams: Improve complex system design by reducing computational cost through automated switching between decomposition-based and non-decomposition-based optimizers.

Authors

Wenjie Qiu, Hepei Li, Jianfeng Wang, Jun Zhang, Yue-Jiao Gong

Abstract

Cooperative Coevolution (CC) and Non-Decomposition Algorithms (NDAs) are two primary approaches for large-scale global optimization (LSGO). Although both have achieved success, they exhibit markedly different performance across problem types and each faces approach-specific limitations, leaving the question of "when to use CC" unresolved. We address this gap with WCC, a novel learning-based framework that dynamically decides when to deploy CC in LSGO. WCC employs a streamlined design for efficient adaptive learning. We formulate the decision as a Markov decision process, designing a targeted feature set, a robust reward, and a tailored network architecture to provide the agent with rich information about the problem state and optimization progress. Our auto-regressive policy enables WCC to perform fine-grained, dynamic switching between CC and NDAs while conditionally allocating the computational budget without expert specifications. WCC shows strong competitiveness and superior generalization across problems, optimization horizons, and baseline optimizers. It realizes a "train-once and reuse" paradigm under a low budget, significantly curbing LSGO cost. Additionally, in-depth analyses yield insights not captured by existing theory, offering practical guidance on when and how to effectively deploy CC.