TABLE III:

Test AUROC for feature incremental learning. The best results are shown in bold.

Methods	Datasets
	CG	CA	DS	AD	CB	BL	IO	IC
LR	0.670	0.773	0.475	0.832	0.727	0.806	0.655	0.825
XGBoost	0.608	0.817	0.527	0.891	0.778	0.816	0.692	0.898
MLP	0.586	0.676	0.516	0.890	0.631	0.825	0.626	0.885
SNN	0.583	0.738	0.442	0.888	0.644	0.818	0.643	0.881
TabNet	0.573	0.689	0.419	0.886	0.571	0.837	0.680	0.882
DCN	0.674	0.835	0.578	0.893	0.778	0.840	0.660	0.891
AutoInt	0.671	0.825	0.563	0.893	0.769	0.836	0.676	0.887
TabTrans	0.653	0.732	0.584	0.856	0.784	0.792	0.674	0.828
FT-Trans	0.662	0.824	0.626	0.892	0.768	0.840	0.645	0.889
VIME	0.621	0.697	0.571	0.892	0.769	0.803	0.683	0.881
SCARF	0.651	0.753	0.556	0.891	0.703	0.829	0.680	0.887
TransTab	0.741	0.879	0.665	0.894	0.791	0.841	0.739	0.897
MambaTab-D	0.787	0.961	0.669	0.904	0.860	0.853	0.783	0.908