Learning-driven Optimization
Overview
Enhancing evolutionary algorithms with machine learning to achieve faster convergence, better generalization, and adaptive behavior in complex optimization scenarios.
Key Questions
- How can learning mechanisms be integrated into evolutionary algorithms without introducing significant computational overhead?
- What types of knowledge should be transferred between optimization phases, and how can negative transfer be avoided?
- How can learned models generalize across different problem instances and dynamic environments?
Our Contributions
- Developed transfer learning frameworks for dynamic multiobjective optimization that reuse historical knowledge to accelerate convergence
- Introduced knee point-based imbalanced transfer learning to handle distribution mismatch in dynamic environments
- Proposed spatial-temporal knowledge transfer for dynamic constrained optimization problems
Recommended Papers
Proposes a transfer learning framework that leverages historical optimization knowledge to accelerate convergence in dynamic multiobjective environments.
Addresses the imbalanced distribution problem in transfer learning for dynamic multiobjective optimization by leveraging knee point information to guide knowledge transfer.
Introduces a spatial-temporal knowledge transfer framework that simultaneously exploits spatial correlations among decision variables and temporal patterns across environmental changes for dynamic constrained multiobjective optimization.
Recent Work
Explores how large language models can serve as intelligent optimizers for dynamic multiobjective problems, leveraging their reasoning and pattern recognition capabilities to guide the evolutionary search.
Open Problems
- Developing theoretically grounded transfer criteria that can determine when and what to transfer
- Scaling learning-driven approaches to high-dimensional and many-objective problems
- Understanding the interaction between learning mechanisms and population diversity
Prospective Student Projects
- Transfer learning for many-objective optimization in dynamic environments
- Meta-learning approaches for rapid adaptation to new optimization problems
- Learning-based constraint handling in dynamic constrained optimization