GeM-EA: A Generative and Meta-learning Enhanced Evolutionary Algorithm for Streaming Data-Driven Optimization

TL;DR

GeM-EA is a novel evolutionary algorithm that combines meta-learning and generative replay to adapt quickly and robustly to streaming data with concept drift.

Contribution

It introduces a unified framework integrating meta-learned surrogate adaptation and generative replay for effective streaming data optimization.

Findings

GeM-EA outperforms existing methods in speed of adaptation.

It demonstrates improved robustness in non-stationary environments.

Experimental results validate the effectiveness of the proposed approach.

Abstract

Streaming Data-Driven Optimization (SDDO) problems arise in many applications where data arrive continuously and the optimization environment evolves over time. Concept drift produces non-stationary landscapes, making optimization methods challenging due to outdated models. Existing approaches often rely on simple surrogate combinations or directly injecting solutions, which may cause negative transfer under sudden environmental changes. We propose GeM-EA, a Generative and Meta-learning Enhanced Evolutionary Algorithm for SDDO that unifies meta-learned surrogate adaptation with generative replay for effective evolutionary search. Upon detecting concept drift, a bi-level meta-learning strategy rapidly initializes the surrogate using environment-relevant priors, while a linear residual component captures global trends. A multi-island evolutionary strategy further leverages historical knowledge via generative replay to accelerate optimization. Experimental results on benchmark SDDO problems demonstrate that GeM-EA achieves faster adaptation and improved robustness compared with state-of-the-art methods.