Degree-Optimized Cumulative Polynomial Kolmogorov-Arnold Networks
This work addresses computational efficiency and data efficiency in neural network design, particularly for regression tasks, but appears incremental as it builds on existing Kolmogorov-Arnold network concepts with a novel optimization approach.
The paper tackles the problem of degree selection in polynomial neural networks by reformulating it as a quadratic unconstrained binary optimization (QUBO) task, reducing complexity from O(D^N) to a single optimization step per layer, and demonstrates competitive performance in regression tasks with limited data, showing good robustness and regularization properties.
We introduce cumulative polynomial Kolmogorov-Arnold networks (CP-KAN), a neural architecture combining Chebyshev polynomial basis functions and quadratic unconstrained binary optimization (QUBO). Our primary contribution involves reformulating the degree selection problem as a QUBO task, reducing the complexity from $O(D^N)$ to a single optimization step per layer. This approach enables efficient degree selection across neurons while maintaining computational tractability. The architecture performs well in regression tasks with limited data, showing good robustness to input scales and natural regularization properties from its polynomial basis. Additionally, theoretical analysis establishes connections between CP-KAN's performance and properties of financial time series. Our empirical validation across multiple domains demonstrates competitive performance compared to several traditional architectures tested, especially in scenarios where data efficiency and numerical stability are important. Our implementation, including strategies for managing computational overhead in larger networks is available in Ref.~\citep{cpkan_implementation}.