AUTHOR=Ma Xiaoliang , Chen Qunjian , Yu Yanan , Sun Yiwen , Ma Lijia , Zhu Zexuan 

TITLE=A Two-Level Transfer Learning Algorithm for Evolutionary Multitasking

JOURNAL=Frontiers in Neuroscience

VOLUME=Volume 13 - 2019

YEAR=2020

URL=https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2019.01408

DOI=10.3389/fnins.2019.01408

ISSN=1662-453X

ABSTRACT=Different from conventional single-task optimization, the recently proposed multitasking optimization (MTO) simultaneously deals with multiple optimization tasks with different types of decision variables. It explores the underlying similarity and complementarity among the component tasks to improve the optimization processes. The well-known multifactorial evolutionary algorithm (MFEA) has been successfully introduced to solve MTO problems based on transfer learning. However, it uses a simple and unstable inter-task transfer learning strategy, thereby resulting in slow convergence. To deal with above issues, this paper presents a two-level transfer learning algorithm, in which the upper-level implements inter-task transfer learning via chromosome crossover and elite individual learning, the lower level introduces intra-task transfer learning based on transferring useful information of decision variables for an across-dimension optimization. The proposed algorithm fully uses the diversity and similarity among component tasks to improve the efficiency and effectiveness of multitasking optimization. Experimental studies demonstrate the proposed algorithm has the outstanding ability of global searching and fast convergence rate