题 目：Learning to Solve Complex Optimisation Problems via Genetic Programming
Many real-world optimisation problems such as vehicle routing, scheduling and resource allocation are very complex, with large problem sizes and dynamic/stochastic environments. Manually designing algorithms to effectively solve these problems requires strong domain expertise and is very time-consuming with many rounds of trial-and-error. This makes it extremely hard to solve the various optimisation problems in the real world.
With the recent rapid progress in AI and machine learning, learning to optimise becomes a hot topic in optimisation, operations research, machine learning and evolutionary computation. To address the difficulty of manually designing effective optimisation algorithms, this trending research area explores the use of machine learning techniques to automatically design algorithms for solving complex optimisation problems. Among all the machine learning techniques, genetic programming is a powerful technique for learning to optimise due to its flexible representation, gradient-free search mechanism to handle the non-differentiable algorithm space, and the inherent advantage in the interpretability of the learned models.
In this presentation, I will introduce how to use genetic programming to learn how to design effective optimisation algorithms. This will cover the fundamental design issues including how to design individual representation, fitness evaluation, parent selection, and genetic operators, as well as more advanced techniques such as feature selection, the use of surrogate models and knowledge transfer. We will use our recent work published on top venues (e.g., IEEE TEVC, TCYB, GECCO) as case studies.
Dr. Yi Mei is an Associate Professor and the Associate Dean (Research) at the Faculty of Engineering, Victoria University of Wellington, New Zealand. He received his BSc and PhD degrees from the University of Science and Technology of China in 2005 and 2010, respectively. His research interests include evolutionary computation for combinatorial optimisation, genetic programming, automatic algorithm design, explainable AI, multi-objective optimisation, transfer/multitask learning and optimisation.
Yi has over 200 fully refereed publications, including the top journals in EC and Operations Research (OR) such as IEEE Transactions on Evolutionary Computation, IEEE Transactions on Cybernetics, European Journal of Operational Research, ACM Transactions on Mathematical Software, and top EC conferences (GECCO). He received an IEEE Transactions on Evolutionary Computation Outstanding Paper Award, two GECCO Best Paper Awards, a GECCO Human Competitive Award, and an EuroGP Best Paper Award. He is an Associate Editor of IEEE Transactions on Evolutionary Computation and an Editorial Board Member of four other international journals. He is the Chair of the IEEE Taskforce on Evolutionary Scheduling and Combinatorial Optimisation and the Chair of the New Zealand Central Section. He is a Fellow of Engineering New Zealand and an IEEE Senior Member.