Abstract:Time series classification is a relevant step supporting decision-making processes in various domains, and deep neural models have shown promising performance. Despite significant advancements in deep learning, the theoretical understanding of how and why complex architectures function remains limited, prompting the need for more interpretable models. Recently, the Kolmogorov-Arnold Networks (KANs) have been proposed as a more interpretable alternative. While KAN-related research is significantly rising, to date, the study of KAN architectures for time series classification has been limited. In this paper, we aim to conduct a comprehensive and robust exploration of the KAN architecture for time series classification on the UCR benchmark. More specifically, we look at a) how reference architectures for forecasting transfer to classification, at the b) hyperparameter and implementation influence on the classification performance in view of finding the one that performs best on the selected benchmark, the c) complexity trade-offs and d) interpretability advantages. Our results show that (1) Efficient KAN outperforms MLP in performance and computational efficiency, showcasing its suitability for tasks classification tasks. (2) Efficient KAN is more stable than KAN across grid sizes, depths, and layer configurations, particularly with lower learning rates. (3) KAN maintains competitive accuracy compared to state-of-the-art models like HIVE-COTE2, with smaller architectures and faster training times, supporting its balance of performance and transparency. (4) The interpretability of the KAN model aligns with findings from SHAP analysis, reinforcing its capacity for transparent decision-making.

Exploring the power of KANs: Overcoming MLP limitations in complex data analysis

KAN: Kolmogorov-Arnold Networks

On the expressiveness and spectral bias of KANs

Kolmogorov-Arnold Networks in Low-Data Regimes: A Comparative Study with Multilayer Perceptrons

Kolmogorov-Arnold Networks (KAN) for Time Series Classification and Robust Analysis

PowerMLP: An Efficient Version of KAN

Can KAN Work? Exploring the Potential of Kolmogorov-Arnold Networks in Computer Vision

Activation Space Selectable Kolmogorov-Arnold Networks

KANs for Computer Vision: An Experimental Study

KAN 2.0: Kolmogorov-Arnold Networks Meet Science

Demonstrating the Efficacy of Kolmogorov-Arnold Networks in Vision Tasks

A Comprehensive Survey on Kolmogorov Arnold Networks (KAN)

Exploring the Limitations of Kolmogorov-Arnold Networks in Classification: Insights to Software Training and Hardware Implementation

KKANs: Kurkova-Kolmogorov-Arnold Networks and Their Learning Dynamics

How to Learn More? Exploring Kolmogorov–Arnold Networks for Hyperspectral Image Classification

A preliminary study on continual learning in computer vision using Kolmogorov-Arnold Networks

A Survey on Kolmogorov-Arnold Network

Exploring Kolmogorov-Arnold Networks for Interpretable Time Series Classification

A Benchmarking Study of Kolmogorov-Arnold Networks on Tabular Data

GraphKAN: Enhancing Feature Extraction with Graph Kolmogorov Arnold Networks

MLP-KAN: Unifying Deep Representation and Function Learning