ICIICII 2020

​​The Fourth International Conference on Industrial Informatics - Computing Technology, Intelligent Technology, Industrial Information Integration 

Shantou, China  

December 18-19, 2020

Invited Speech

​​

Local landscape

​​

Co-sponsors

Kangshun Li, PhD,Professor


华南农业大学




Title: 基于适应度景观的差分进化算法变异算子及参数选择研究 


Abstract

      为了克服传统差分算法评估待求解问题适应值时计算资源需求大且效率低的问题,该研究提取经典基准函数的适应度景观特征并通过一系列实验得出对应特征适合的变异算子及算法参数设置,采用机器学习方法建立适应度景观与变异算子以及算法参数的关系;使用通过学习获得的分类器在求解复杂问题时采样少量的点并计算问题的适应度景观特征得到合适的变异算子和算法参数。首先分析DE/best/1DE/current-to-rand/1两种变异算子在不同类型问题上的性能差异;然后采用集成学习与决策树建立问题的适应度景观与变异算子之间的联系,得到名为变异算子选择器的分类器;接着采用神经网络方法建立问题的适应度景观与算法参数之间的联系,得到名为参数选择器的分类器。将训练得到的分类器在CEC2017问题集上进行测试,根据问题的适应度景观为差分进化算法选择变异算子和算法参数,结果表明,采用这种选择算子与设置算法参数策略的平均性能比使用单一算子的平均性能好。


Biography
     李康顺,教授,博士后,博士生导师,江西兴国人;福建省闽江学者,IEEE Guangzhou Computational Intelligence Society (CIS) 副主席、中国仿真学会智能仿真优化与调度专委会常务委员,广东省计算智能专业委员会主任,IEEE高级会员、国家重点研发项目评审专家。江西大学(现南昌大学)计算数学专业本科毕业并获理学学士学位,武汉大学计算机软件与理论专业博士毕业并获工学博士学位,中国科学院自动化研究所控制科学和工程专业博士后流动站出站,加拿大University of Calgary访问教授,香港中文大学访问教授;历任江西省赣州地区统计局计算中心主任、高级工程师,江西理工大学、华南农业大学教授、博士生导师、院长,江西预备役师少校;省级新世纪百千万人才、省级中青年学科带头人;Information Sciences (SCI一区) 副主编、International Journal of Cognitive Informatics and Natural Intelligence (IJCINI)(EI收录) 主编、AIMS Agriculture and Food-Open Access Journals (https://www.aimspress.com/journal/agriculture)编委、《系统仿真学报》编委、《江西理工大学学报》编委。发表高水平论文152篇,其中:SCI收录31篇、EI收录106篇;出版英文论著3部。主持国家自然科学基金2项、国家统计科研重点项目1项、国家重大专项子课题1项、中国博士后重点基金项目1项,主持广东省重点研发项目1项、省自然科学基金3项、省攻关项目6项,获省级统计科研成果三等奖1项,获省教育厅科技成果三等奖1项,指导学生获得“挑战杯”广东大学生创业大赛创业实践挑战赛银奖,获省级双语教学示范课程1项。主要从事人工智能、机器学习、图像识别、工业互联网、进化计算、大数据建模和神经网络等理论与应用研究。担任过四个国际会议的主席.
  • The bird's-eye view of Shantou University
  • Nan’ao  Island
  • Nan'ao Bridge
  • Zhongshan Pavilion
  • Shantou Coastal Corridor
  • Lotus Pond
  • Gentleman Sculpture Group
Ke Tang, PhD,Professor


Department of Computer Science and Engineering, 

Southern University of Science and Technology



Title: Scalable Evolutionary Search



Abstract
     In the past decades, Evolutionary Algorithms (EAs) have been shown to be capable of achieving high quality solutions to many hard optimization problems (e.g., NP-Hard problems). On the other hand, it is often criticized as having high computational cost, especially for large-scale problems. This drawback is becoming a major hurdle for EAs’ applications in the real-world. This talk will present advances on making EAs more scalable in 3 most common circumstances, which aim at accelerating EAs significantly without deteriorate the quality of obtained solutions. In addition to general ideas and methodologies, case studies on vehicle routing, deep learning and social network analysis will also be given.

Biography
    Ke Tang is a Professor at the Department of Computer Science and Engineering, Southern University of Science and Technology (SUSTech). Before that, he was with University of Science and Technology of China, first as an Associate Professor (2007-2011) and then a full Professor (2011-2017). His major research interests include evolutionary computation, machine learning and their applications. He has published more than 160 journal and conference papers. He is/was an Associate Editor or Editorial Board Member of the IEEE Trans. on Evolutionary Computation, IEEE Computational Intelligence Magazine, Computational Optimization and Applications (Springer), Natural Computing (Springer) and Memetic Computing (Springer) and served as program/technical chairs/co-chairs of 10 international conferences. He received the Royal Society Newton Advanced Fellowship in 2015 and the 2018 IEEE Computational Intelligence Society Outstanding Early Career Award.
Next
Previous
First
Last