任务间共享和特有结构分解的多任务TSK模糊系统建模

PDF下载HTML

Title:

Multi-task TSK fuzzy system modeling based on inter-task common and special structure decomposition

作者:

赵壮壮¹, 王骏², 潘祥¹, 邓赵红¹, 施俊², 王士同¹

1. 江南大学 人工智能与计算机学院，江苏 无锡 214122;
2. 上海大学 通信与信息工程学院，上海 200444

Author(s):

ZHAO Zhuangzhuang¹, WANG Jun², PAN Xiang¹, DENG Zhaohong¹, SHI Jun², WANG Shitong¹

1. School of Artificial Intelligence and Computer Science, Jiangnan University, Wuxi 214122, China;
2. School of Communication & Information Engineering, Shanghai University, Shanghai 200444, China

关键词:

TSK模糊系统; 非线性; 多任务; 低秩; 稀疏; 参数分解; 泛化性能; 可解释性

Keywords:

TSK fuzzy system; nonlinear; multitask; low-rank; sparse; parameter decomposition; generalization performance; interpretability

分类号:

TP181

DOI:

10.11992/tis.202007009

摘要:

现有的多任务Takagi-Sugeno-Kang (TSK) 模糊建模方法更注重利用任务间的相关性信息，而忽略了单个任务的特殊性。针对此问题，本文提出了一种考虑所有任务之间的共享结构和特有结构的TSK模糊系统多任务建模新方法。该方法将后件参数分解为共享参数和特有参数两个分量，既充分利用了任务间共享信息，又有效地保留了单个任务的特性。最后，本文利用增广拉格朗日乘子法(ALM)求解该最优化问题。实验结果表明，该方法比现有的模型获得了更好的表现。

Abstract:

Existing Takagi–Sugeno–Kang (TSK) fuzzy system modeling methods pay more attention to the inter-task correlation but ignore the particularity of every single task. To address this issue, this paper proposes a novel multi-task modeling method for TSK fuzzy systems taking common and specific structures across all tasks (MTTSKFS-CS) into consideration. This method decomposes consequent parameters into shared and special ones, which not only takes advantage of the shared information among tasks but also effectively preserves the characteristics of individual tasks. Finally, the study uses the augmented Lagrange multiplier for optimization. The experimental results demonstrate the better performance of the proposed model compared with other existing methods.