[Report] How to use Autoencoder and Surrogate Model to Accelerate Evolutionary Algorithms

July 10 09:00-11:00, 2023

Editor:College of Information Science and Technology Time:2023-07-09


Speaker: MengChu Zhou

Time: July 10 09:00-11:00, 2023

Venue: The second lecture hall of Science Hall


Abstract:High-dimensional computationally expensive problems (HEPs) in which a single fitness evaluation consumes hours or even days have attracted much attention from both academia and industry. Exponentially expanding search space and complex landscape brought by numerous decision variables make HEPs extremely challenging to be solved by traditional algorithms with limited physical/computational resources. Therefore, an Autoencoder-embedded Evolutionary Optimization (AEO) framework is invented to deal with them. To be specific, high-dimensional search space can be compressed to informative low-dimensional space by using an autoencoder as an effective dimension reduction tool. The search operation conducted in this low-dimensional space facilitates the population in convergence towards the optima. To balance the exploration and exploitation ability during optimization, two sub-populations are adopted to coevolve in a distributed/parallel fashion, wherein one is assisted by an autoencoder and the other undergoes a regular evolutionary process. Dynamic information exchange is conducted between them after each cycle to promote population diversity. Moreover, surrogate models can be incorporated into AEO (SAEO) to further boost its performance by reducing unnecessary fitness evaluation. Compared with the state-of-the-art algorithms for HEPs, AEO shows extraordinarily high efficiency for these challenging problems while SAEO can greatly improve the performance of AEO in most cases, thus opening new directions for various swarm optimization and evolutionary algorithms under both AEO and SAEO to tackle HEPs and greatly advancing the field of high-dimensional computationally expensive optimization.


Biography: MengChu Zhou received his B.S. degree in Control Engineering from Nanjing University of Science and Technology, Nanjing, China in 1983, M.S. degree in Automatic Control from Beijing Institute of Technology, Beijing, China in 1986, and Ph. D. degree in Computer and Systems Engineering from Rensselaer Polytechnic Institute, Troy, NY in 1990.  He joined New Jersey Institute of Technology (NJIT), Newark, NJ in 1990, and has been Distinguished Professor in Electrical and Computer Engineering since 2013. His research interests are in Petri nets, intelligent automation, Cloud/edge Computing, Internet of Things, big data, web services, and intelligent transportation.  He has over 1100 publications including 14 books, 750+ journal papers (600+ in IEEE transactions), 31patents and 32 book-chapters. He is the founding Editor of IEEE Press Book Series on Systems Science and Engineering, and Associate Editor of IEEE Internet of Things Journal, IEEE Transactions on Intelligent Transportation Systems, and IEEE Transactions on Systems, Man, and Cybernetics: Systems. He was Editor-in-Chief of IEEE/CAA Journal of Automatica Sinica (2018-2022). He is a recipient of Humboldt Research Award for US Senior Scientists from Alexander von Humboldt Foundation, Franklin V. Taylor Memorial Award and the Norbert Wiener Award from IEEE Systems, Man and Cybernetics Society, Excellence in Research Prize and Medal from NJIT, and Edison Patent Award from the Research & Development Council of New Jersey. He is highly cited scholar with over 60,500 Google Scholar citations and h-index 123. He has been among most highly cited scholars since 2012 and ranked top one in the field of engineering worldwide in 2012 by Web of Science. He has over 58,600 GoogleScholar citations with h-index 122. He was ranked #89 in the world and #58 in the United States among the 2022 Top 1000 Scientists in Computer Science in the World, Research.com. He is a life member of Chinese Association for Science and Technology-USA and served as its President in 1999. He is a Fellow of IEEE, International Federation of Automatic Control (IFAC), American Association for the Advancement of Science (AAAS), Chinese Association of Automation (CAA), and National Academy of Inventors (NAI).