Table 3.

Comparison of model fit under different metrics

Model type	Accuracy	AUC	AIC
Deontological	0.630	0.631	1.303
Utilitarian	0.719	0.779	1.161
Hybrid	0.756	0.814	1.052
Hybrid + Axes (iteration 1)	0.760	0.823	1.021
Additional Axes (iteration 2)	0.764	0.825	1.019
Two-way conjunctions (iteration 3)	0.764	0.829	1.003
Three-way conjunctions (iteration 4)	0.768	0.830	0.999
Neural network	0.768	0.833	0.999
Empirical upper bound	0.804	0.890	N/A

The AIC requires the number of model parameters, which is not applicable (N/A) for the empirical upper bound.