![](/__local/4/91/8C/BFAECFB7E99AB79964932547BEE_2DA0B431_1BF90.jpg)
报告主题:
Evolutionary Multi-objective Optimization for Practical Problem-Solving Tasks
报告时间:
2024年5月24日10:00
报告地点:
人工智能与计算机学院B310会议室
报告摘要:
Most practical problem-solving tasks are better posed with multiple conflicting objectives, such as, simultaneous optimization of cost and qu3lity, rather than with a single aggregated criterion. Multi-objective optimization problems give rise to a set of trade-off Pareto-optimal solutions, which must first be found and then a single preferred solution must be chosen for implementation. This requires a coordination of computational techniques executing optimization and decision-making. In this talk, we briefly discuss four uses of evolutionary multi-objective optimization (EMO) methods, including a knowledge discovery task which is getting a lot of attention. We shall also highlight a few current research areas on the topic, including the use of machine learning methods to improve EMO's performance, and demonstrate their usefulness through examples of real-world problems.
主讲人:
Kalyanmoy Deb,美国密歇根州立大学电子计算工程学院首席教授,印度国家工程学院院士和国家科学院院士,分别于2012年、2022年当选IEEE Fellow、ACM Fellow。Deb教授主要从事进化优化算法及其在多目标优化、建模和机器学习领域的应用研究。Deb教授已发表文章620余篇,谷歌他引超20万次,h-指数139,长期担任进化计算领域顶级期刊Evolutionary Computation Journal (ECJ)以及工程与优化领域相关10多个期刊的咨询委员会成员,受邀在重大国际会议和专题研讨会上发表了120余次主题演讲。Deb教授获得科睿唯安终身成就奖、国际软计算学会杰出科学家奖、IEEE CIS进化计算先锋奖、工程科学TWAS奖等。