Evolutionary Computation (EC) has been widely applied to many learning and optimisation problems, such as clustering, combinatorial optimisation, classification, regression, and clustering. Its population-based search mechanism addresses the problems by exploring and evaluating many candidate solutions, which discovers much knowledge about the problems. However, most existing evolutionary learning and optimisation paradigms start from scratch when solving a new problem, despite the similarity between the new problem and the addressed problems. However, utilising the knowledge learned from the addressed problems can significantly improve the performance and convergence speed on other related problems, which is the main idea of transfer learning. Transfer learning is an emergent and hot topic in data mining and machine learning. Recently, transfer learning has been gained much attention by the EC community to address challenging learning and optimisation issues.
The theme of this special session is evolutionary transfer learning, covering both new theories and methods to transfer knowledge with different EC paradigms, how transfer learning can be adopted in EC, and the applications of evolutionary transfer learning in real-world problems. The EC paradigms include but not limited to Genetic Programming (GP), Genetic Algorithms (GAs), Particle Swarm Optimisation (PSO), Differential Evolution (DE), Evolutionary Strategy (ES), Ant Colony Optimisation (ACO), and Evolutionary Multi-objective Optimisation (EMO).
Transfer learning and domain adaptation are interdisciplinary topics that are related to several symposia such as IEEE Symposium on Evolutionary Scheduling and Combinatorial Optimisation, IEEE Symposium on Computational Intelligence in Feature Analysis, Selection and Learning in Image and Pattern Recognition (IEEE FASLIP), IEEE Symposium on Deep Learning (IEEE DL). However, none of the above symposia explicitly covers transfer learning and domain adaptation.
Authors are invited to submit their original and unpublished work to this special session. Topics of interest include but are not limited to:
- Evolutionary supervised and unsupervised transfer learning
- Evolutionary deep transfer learning
- Evolutionary domain adaptation and domain generalisation
- Instance-based transfer learning in Evolutionary Computation
- Feature-based transfer learning in Evolutionary Computation
- Evolutionary transfer learning for clustering, combinatorial optimisation, classification, regression, and association rules
- Evolutionary transfer learning for scheduling and combinatorial optimisation
- Evolutionary transfer learning for multi-objective and many-objective optimisation
- Evolutionary multi-task optimisation
- Theoretical studies on the behaviours of evolutionary transfer learning andoptimisation
- Evolutionary transfer learning for real-world applications such as computervision, image analysis, cybersecurity
Organizers: Bach Nguyen, Liang Feng, Bing Xue, Mengjie Zhang