Kolmogorov Arnold Networks in Fraud Detection: Bridging the Gap Between Theory and Practice

TL;DR

A quick decision rule is proposed using Principal Component Analysis to assess the suitability of KAN: if data can be effectively separated in two dimensions using splines, KAN may outperform traditional models; otherwise, other methods could be more appropriate.

Abstract

This study evaluates the applicability of Kolmogorov-Arnold Networks (KAN) in fraud detection, finding that their effectiveness is context-dependent. We propose a quick decision rule using Principal Component Analysis (PCA) to assess the suitability of KAN: if data can be effectively separated in two dimensions using splines, KAN may outperform traditional models; otherwise, other methods could be more appropriate. We also introduce a heuristic approach to hyperparameter tuning, significantly reducing computational costs. These findings suggest that while KAN has potential, its use should be guided by data-specific assessments.

Figures (8)

• Figure 1: KAN Confusion Matrix Using Genetic Algorithm for Dataset 1
• Figure 2: PCA Results for Dataset 1
• Figure 3: PCA Results for Datasets 2-5
• Figure 4: KAN Confusion Matrix for Dataset 1
• Figure 5: KAN Confusion Matrix for Dataset 2