报告题目：Evolutionary Computation and Learning for CombinatorialOptimisation: 15 Years of My Experience and Reflections
报告人：Yi Mei， Victoria University of Wellington
报告时间：2023年7月21日14:00-15:00
报告地点：海山楼二楼会议室（B0204）
内容简介：
Many real-world problems such as vehicle routing scheduling and resource allocation are complexcombinatorial optimisation problems with various challenges like large problem sizes and dynamic/stochasticenvironments. Most such problems are known as NP-hard and cannot be completely solved in reasonable timeEvolutionaryComputation(EC) hasbeen successful in solving combinatorial optimisation problemsfinding near-optimal solutions in reasonable time. Compared with exact methods (e.g.approximatelymathematical programming). EC is much faster. Compared with other heuristic and meta-heuristic method.(e.g.. simulated annealing), EC can jump out of local optima more easily. Overall. EC offers a better balancebetween exploration and exploitation in the research spaceIn this talk, I will share my experience after working in this area for over 15 years. It consists of three marparts, indicating how my research nterests have been shifted over time throughout the years. The first part is ECfor classic static combinatorial optimisation, including the design of representation and search operators, dealingwith large scale problem sizes. The second part is EC for dynamic stochastic combinatorial optimisation. mainlypromising decision-making rules that can responofocused on using Genetic Programming (GP) to learn/evolveto the dynamic environment in real time. The third part is EC for automatic algorithm design. switching frommanually designing algorithms to solve specific problem instances to leveraging machine leaming paradigm tcautomatically design effective algorithms. This is also known as a recenthot cross-disciplinarv research areacalled “learn to optimise". In each part, I will introduce some exemplar works that our group has done. In theend. I will share the lessons I learned and reflections I have from so many vears of research in this area.
报告人简介：
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 ofScience and Technology of China in 2005 and 2010, respectively. His research interests inchude evolutionarycomputation for combinatorial optimisation, genetic programming, automatic algorithm design, explainable Almulti-objective optimisation, transfer/multitask learning and optimisation.Yi has about 200 fully refereed publications, including the top journals im EC and Operations Research (ORsuch as IEEE Transactions on Evolutionary Computation, IEEE Transactions on Cyberneics. European Journaof Operational Research, ACM Transactions on Mathematical Software, and top EC conferences (GECCO). Hewon an IEEE Transactions on Evolutionary Computation Outstanding Paper Award 2017, a GECCO Best PapeiAward 2022 and the EuroGP Best Paper Award 2022. He is an Associate Editor of IEEE Transactions orEvolutionary Computation and an Editorial Board Member of four other international joumals. He serves as aVice-Chair of the IEEE CIS Emergent Technologies Technical Committee and a member of three IEEE CISTask Forces and two IEEE CIS Technical Committees. He is the Chair of the IEEE Taskforce on EvolutionaryScheduling and Combinatoral Optimisation and the Char of the New Zealand Central Section. He is a FellowofEngineering New Zealand and an IEEE Senior Member.
