Learning large-scale fuzzy cognitive maps using an evolutionary many-task algorithm

• Published in: Applied soft computing Vol. 108; p. 107441
• Main Authors: Wang, Chao, Liu, Jing, Wu, Kai, Ying, Chaolong
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.09.2021
• Subjects: Batch learning; Evolutionary algorithm; Fuzzy cognitive maps (FCMs); Many-task optimization; Multi-task optimization; Random inactivation; Fuzzy cognitive maps (FCMs); Random inactivation; Many-task optimization; Multi-task optimization; Evolutionary algorithm; Batch learning
• ISSN: 1568-4946; 1872-9681
• DOI: 10.1016/j.asoc.2021.107441

Fuzzy cognitive maps (FCMs) are a powerful tool for simulating and analyzing complex systems. Many efficient methods based on evolutionary algorithms have been proposed to learn small-scale FCMs. However, large number of function evaluations of those methods make them difficult to cope with large-scale FCM learning problems. To overcome this issue, we propose a random inactivation-based batch many-task evolutionary algorithm, termed as IBMTEA-FCM. Inspired by the probability of knowledge sharing in different tasks, the problem of FCM learning is first modeled as a many-task optimization problem, in which each task represents learning local connections of a node in a single FCM. To ensure the effectiveness of knowledge transfer, all tasks are randomly divided into multiple batches to optimize separately. In this method, an evolutionary many-task framework is employed to overcome the proposed many-task FCM learning problem and we randomly deactivate weighted edges to ensure the sparsity of FCM in the evolutionary process. The performance of IBMTEA-FCM is validated on both synthetic datasets and a practical study of gene regulatory network reconstruction. Compared with existing classical methods, the experimental results show that IBMTEA-FCM can learn large-scale FCMs with higher accuracy and less computational cost. •Propose a batch evolutionary many-task framework to overcome the many-task FCM learning problem.•Optimize multiple tasks in the same batch by the knowledge sharing between different tasks.•Random inactivation is designed to reduce the risk of the FCM model overfitting.•Our method depicts a superior capability in gene regulatory network reconstruction.