Chaotic Time Series Prediction Using Neuro-Fuzzy Systems with Cluster-Based Tribes Optimization Algorithm

Abstract

This study presents an efficient cluster-based tribes optimization algorithm (CTOA) to design neuro-fuzzy systems (NFS) for chaotic time series prediction. The proposed CTOA learning algorithm was used to parameter optimization of the NFS model. The CTOA adopts a self-clustering algorithm (SCA) to divide suitably a swarm into multiple tribes and uses different displacement strategies let each particle to select to update. Furthermore, the CTOA also utilizes adaptation mechanism to generate or remove particles and reconstruct tribal links to make the tribes to more adaption and improve the qualities of the tribes to evolve. Finally, the proposed NFS-CTOA method is applied to predict chaotic time series. Results of this study demonstrate the effectiveness of the proposed CTOA learning algorithm.