Learning Evolution Via Optimization Knowledge Adaptation
- 23 April 2026
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. PP (01628828) , 1-18
Abstract
The iterative search process of evolutionary algorithms (EAs) encapsulates optimization knowledge within historical populations and fitness evaluations. Effective utilization of this knowledge is crucial for facilitating knowledge transfer and online adaptation. However, current research typically addresses these goals in isolation and faces distinct limitations: evolutionary sequential transfer optimization often suffers from incomplete utilization of prior knowledge, while adaptive strategies, utilizing real-time knowledge, are limited to tailoring specific evolutionary operators. To simultaneously achieve these two capabilities, we introduce the Optimization Knowledge Adaptation Evolutionary Model (OKAEM), a unified learnable evolutionary framework capable of adaptively updating parameters based on available optimization knowledge. By parameterizing evolutionary operators via attention mechanisms, OKAEM enables learnable update rules that facilitate the utilization of optimization knowledge via two phases: pre-training to integrate extensive prior knowledge for efficient transfer, and adaptive optimization to dynamically update parameters based on real-time knowledge. Experimental results confirm that OKAEM significantly outperforms state-of-the-art sequential transfer methods across 12 transfer scenarios via pre-training, and surpasses advanced learnable EAs solely through its self-tuning mechanism in prior-free settings. Beyond demonstrating practical utility in prompt tuning for vision-language models, ablation studies validate the necessity of the learnable components, while visualization analyses reveal the model's capacity to autonomously discover interpretable evolutionary principles.Keywords
Funding Information
- National Natural Science Foundation of China (U22B2054)
- National Natural Science Foundation of China (62076192)
- Higher Education Discipline Innovation Project
- Program for Cheung Kong Scholars and Innovative Research Team in University (IRT 15R53)
- Science and Technology Innovation Project from the Chinese Ministry of Education
- National Key Laboratory of Human-Machine Hybrid Augmented Intelligence, Xi’an Jiaotong University (HMHAI-202404)
This publication has 67 references indexed in Scilit:
- Evolution Through Large Models2023
- Learning to Prompt for Vision-Language ModelsInternational Journal of Computer Vision, 2022
- Evolutionary multi-task optimization with hybrid knowledge transfer strategyInformation Sciences, 2021
- Learning Step-Size Adaptation in CMA-ES2020
- Meta-Learning for Black-Box Optimization2020
- Evolutionary Learning: Advances in Theories and Algorithms2019
- Food-101 – Mining Discriminative Components with Random Forests2014
- Computing Machinery and Intelligence2007
- The CMA Evolution Strategy: A Comparing Review2007
- Evolution strategies: A family of non-linear optimization techniques based on imitating some principles of organic evolutionAnnals of Operations Research, 1984