Improving Generalizability of Kolmogorov-Arnold Networks via Error-Correcting Output Codes
This work addresses the problem of enhancing KAN performance for multi-class medical image classification in healthcare AI, though it appears incremental as it applies an existing technique (ECOC) to a new model (KAN).
The paper tackled improving the generalizability of Kolmogorov-Arnold Networks (KAN) by integrating Error-Correcting Output Codes (ECOC) for multi-class classification, resulting in higher accuracy on a blood cell classification dataset across various hyperparameter settings.
Kolmogorov-Arnold Networks (KAN) offer universal function approximation using univariate spline compositions without nonlinear activations. In this work, we integrate Error-Correcting Output Codes (ECOC) into the KAN framework to transform multi-class classification into multiple binary tasks, improving robustness via Hamming distance decoding. Our proposed KAN with ECOC framework outperforms vanilla KAN on a challenging blood cell classification dataset, achieving higher accuracy across diverse hyperparameter settings. Ablation studies further confirm that ECOC consistently enhances performance across FastKAN and FasterKAN variants. These results demonstrate that ECOC integration significantly boosts KAN generalizability in critical healthcare AI applications. To the best of our knowledge, this is the first work of ECOC with KAN for enhancing multi-class medical image classification performance.