Hierarchical RBF-KAN and RBF-SKAN architectures for function approximation

Arxiv CS.LG (machine learning)

Proposes hierarchical Kolmogorov-Arnold network architectures using radial basis functions as activations for deterministic and random field approximation. Includes theoretical analysis and numerical experiments.

Arxiv CS.LG (machine learning)·Mingtao XiaQijing Shen

8 days ago