LTBs-KAN: Linear-Time B-splines Kolmogorov-Arnold Networks
This work addresses the computational bottleneck of KANs, making them more practical for researchers and practitioners interested in explainable neural networks.
LTBs-KAN proposes a linear-time B-spline computation method for Kolmogorov-Arnold Networks, reducing computational complexity and parameter count. On MNIST, Fashion-MNIST, and CIFAR-10, it achieves faster training and comparable accuracy to other KAN implementations.
Kolmogorov-Arnold Networks (KANs) are a recent neural network architecture offering an alternative to Multilayer Perceptrons (MLPs) with improved explainability and expressibility. However, KANs are significantly slower than MLPs due to the recursive nature of B-spline function computations, limiting their application. This work addresses these issues by proposing a novel base-spline Linear-Time B-splines Kolmogorov-Arnold Network (LTBs-KAN) with linear complexity. Unlike previous methods that rely on the Boor-Mansfield-Cox spline algorithm or other computationally intensive mathematical functions, our approach significantly reduces the computational burden. Additionally, we further reduce model's parameter through product-of-sums matrix factorization in the forward pass without sacrificing performance. Experiments on MNIST, Fashion-MNIST and CIFAR-10 demonstrate that LTBs-KAN achieves good time complexity and parameter reduction, when used as building architectural blocks, compared to other KAN implementations.