Kolmogorov Arnold Networks (KANs) for Imbalanced Data -- An Empirical Perspective

TL;DR

This paper empirically evaluates Kolmogorov Arnold Networks (KANs) for imbalanced classification, revealing their strengths in raw data scenarios but highlighting significant computational costs and conflicts with standard imbalance techniques.

Contribution

It provides the first comprehensive empirical assessment of KANs in imbalanced data contexts, identifying their limitations and guiding future architectural improvements.

Findings

KANs outperform MLPs on raw imbalanced data without resampling.

Resampling and focal loss degrade KAN performance significantly.

MLPs with imbalance techniques match KANs at lower resource costs.

Abstract

Kolmogorov Arnold Networks (KANs) are recent architectural advancement in neural computation that offer a mathematically grounded alternative to standard neural networks. This study presents an empirical evaluation of KANs in context of class imbalanced classification, using ten benchmark datasets. We observe that KANs can inherently perform well on raw imbalanced data more effectively than Multi-Layer Perceptrons (MLPs) without any resampling strategy. However, conventional imbalance strategies fundamentally conflict with KANs mathematical structure as resampling and focal loss implementations significantly degrade KANs performance, while marginally benefiting MLPs. Crucially, KANs suffer from prohibitive computational costs without proportional performance gains. Statistical validation confirms that MLPs with imbalance techniques achieve equivalence with KANs (|d| < 0.08 across metrics) at minimal resource costs. These findings reveal that KANs represent a specialized solution for raw imbalanced data where resources permit. But their severe performance-resource tradeoffs and incompatibility with standard resampling techniques currently limits practical deployment. We identify critical research priorities as developing KAN specific architectural modifications for imbalance learning, optimizing computational efficiency, and theoretical reconciling their conflict with data augmentation. This work establishes foundational insights for next generation KAN architectures in imbalanced classification scenarios.