Clifford Kolmogorov-Arnold 网络（ClKAN）提出
Clifford Kolmogorov-Arnold Networks
AI 导读
研究者提出 Clifford Kolmogorov-Arnold 网络（ClKAN），一种在任意 Clifford 代数空间中进行函数逼近的灵活高效架构。针对高维代数的指数级扩展问题，其采用 Randomized Quasi-Monte Carlo 网格生成方案，并引入新的批归一化策略处理变域输入。ClKAN 在合成任务和物理启发任务上完成验证，可应用于科学发现与工程领域。
正文 · 原文
Abstract:We introduce Clifford Kolmogorov-Arnold Network (ClKAN), a flexible and efficient architecture for function approximation in arbitrary Clifford Algebra spaces. We propose the use of Randomized Quasi-Monte Carlo grid generation as a solution to the exponential scaling associated with higher-dimensional algebras. Our ClKAN also introduces new batch normalization strategies to deal with variable domain input. ClKAN finds application in scientific discovery and engineering, and is validated in synthetic and physics-inspired tasks.
| Comments: | published in proceedings of IEEE International Joint Conference on Neural Networks (IJCNN) 2026 |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2602.05977 [cs.LG] |
| (or arXiv:2602.05977v3 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2602.05977 arXiv-issued DOI via DataCite |
|
| Related DOI: | https://doi.org/10.1109/IJCNN69854.2026.11728617
DOI(s) linking to related resources |
Submission history
From: Matthias Wolff [view email]
[v1]
Thu, 5 Feb 2026 18:25:40 UTC (315 KB)
[v2]
Tue, 23 Jun 2026 22:02:43 UTC (321 KB)
[v3]
Fri, 23 Oct 2026 14:03:55 UTC (323 KB)
来源：arXiv cs.AI · arxiv.org