DP-KAN: Differentially Private Kolmogorov-Arnold Networks
This work addresses privacy-preserving training for KANs, an incremental extension of existing methods to a new model type.
The paper tackled the problem of applying differential privacy to Kolmogorov-Arnold Networks (KANs) using DP-SGD, finding that KANs achieve accuracy comparable to MLPs and show similar degradation under privacy constraints.
We study the Kolmogorov-Arnold Network (KAN), recently proposed as an alternative to the classical Multilayer Perceptron (MLP), in the application for differentially private model training. Using the DP-SGD algorithm, we demonstrate that KAN can be made private in a straightforward manner and evaluated its performance across several datasets. Our results indicate that the accuracy of KAN is not only comparable with MLP but also experiences similar deterioration due to privacy constraints, making it suitable for differentially private model training.