Fig. 4: AUC performance metrics across models and datasets. | Communications Medicine

Fig. 4: AUC performance metrics across models and datasets.

From: Volumetric spline-based Kolmogorov-Arnold architectures surpass CNNs, vision transformers, and graph networks for Parkinson’s disease detection

Fig. 4: AUC performance metrics across models and datasets.

a Isolated dataset analysis showing individual dataset performance. b Combined dataset analysis with merged datasets. c Hold-out analysis for cross-dataset generalization. Heatmap colors represent AUC values ranging from 0.4 (dark blue) to 1.0 (dark red); n = 59 independent subjects (PPMI), n = 43 (NEUROCON), n = 40 (Tao Wu) for analyses.

Back to article page