Time Series Forecasting with Hahn Kolmogorov-Arnold Networks

Md Zahidul Hasan, Abdessamad Ben Hamza, Nizar Bouguila
Proceedings of The 29th International Conference on Artificial Intelligence and Statistics, PMLR 300:2188-2196, 2026.

Abstract

Recent Transformer- and MLP-based models have demonstrated strong performance in long-term time series forecasting, yet Transformers remain limited by their quadratic complexity and permutation-equivariant attention, while MLPs exhibit spectral bias. We propose HaKAN, a versatile model based on Kolmogorov-Arnold Networks (KANs), leveraging Hahn polynomial-based learnable activation functions and providing a lightweight and interpretable alternative for multivariate time series forecasting. Our model integrates channel independence, patching, a stack of Hahn-KAN blocks with residual connections, and a bottleneck structure comprised of two fully connected layers. The Hahn-KAN block consists of inter- and intra-patch KAN layers to effectively capture both global and local temporal patterns. Extensive experiments on various forecasting benchmarks demonstrate that our model consistently outperforms recent state-of-the-art methods, with ablation studies validating the effectiveness of its core components.

Cite this Paper

BibTeX
@InProceedings{pmlr-v300-hasan26a, title = { Time Series Forecasting with Hahn Kolmogorov-Arnold Networks }, author = {Hasan, Md Zahidul and Hamza, Abdessamad Ben and Bouguila, Nizar}, booktitle = {Proceedings of The 29th International Conference on Artificial Intelligence and Statistics}, pages = {2188--2196}, year = {2026}, editor = {Khan, Emtiyaz and Li, Yingzhen and Solin, Arno and Ramdas, Aaditya}, volume = {300}, series = {Proceedings of Machine Learning Research}, month = {02--05 May}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v300/main/assets/hasan26a/hasan26a.pdf}, url = {https://proceedings.mlr.press/v300/hasan26a.html}, abstract = { Recent Transformer- and MLP-based models have demonstrated strong performance in long-term time series forecasting, yet Transformers remain limited by their quadratic complexity and permutation-equivariant attention, while MLPs exhibit spectral bias. We propose HaKAN, a versatile model based on Kolmogorov-Arnold Networks (KANs), leveraging Hahn polynomial-based learnable activation functions and providing a lightweight and interpretable alternative for multivariate time series forecasting. Our model integrates channel independence, patching, a stack of Hahn-KAN blocks with residual connections, and a bottleneck structure comprised of two fully connected layers. The Hahn-KAN block consists of inter- and intra-patch KAN layers to effectively capture both global and local temporal patterns. Extensive experiments on various forecasting benchmarks demonstrate that our model consistently outperforms recent state-of-the-art methods, with ablation studies validating the effectiveness of its core components. } }
Endnote
%0 Conference Paper %T Time Series Forecasting with Hahn Kolmogorov-Arnold Networks %A Md Zahidul Hasan %A Abdessamad Ben Hamza %A Nizar Bouguila %B Proceedings of The 29th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2026 %E Emtiyaz Khan %E Yingzhen Li %E Arno Solin %E Aaditya Ramdas %F pmlr-v300-hasan26a %I PMLR %P 2188--2196 %U https://proceedings.mlr.press/v300/hasan26a.html %V 300 %X Recent Transformer- and MLP-based models have demonstrated strong performance in long-term time series forecasting, yet Transformers remain limited by their quadratic complexity and permutation-equivariant attention, while MLPs exhibit spectral bias. We propose HaKAN, a versatile model based on Kolmogorov-Arnold Networks (KANs), leveraging Hahn polynomial-based learnable activation functions and providing a lightweight and interpretable alternative for multivariate time series forecasting. Our model integrates channel independence, patching, a stack of Hahn-KAN blocks with residual connections, and a bottleneck structure comprised of two fully connected layers. The Hahn-KAN block consists of inter- and intra-patch KAN layers to effectively capture both global and local temporal patterns. Extensive experiments on various forecasting benchmarks demonstrate that our model consistently outperforms recent state-of-the-art methods, with ablation studies validating the effectiveness of its core components.
APA
Hasan, M.Z., Hamza, A.B. & Bouguila, N.. (2026). Time Series Forecasting with Hahn Kolmogorov-Arnold Networks . Proceedings of The 29th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 300:2188-2196 Available from https://proceedings.mlr.press/v300/hasan26a.html.

Related Material