Advancing Preference Learning in AI: Beyond Pairwise Comparisons

Table 1.

Advancing Preference Learning in AI: BeyondPairwise Comparisons.


Study	Method	Reduction in Accuracy Maintained (%)	Comparisons (%)

Qian et al. (2015)	Adaptive PairwiseComparison	30%	95%
Ignatenko et al. (2021)	Bayesian Optimization Framework	35%	93%

^a Results of Pairwise Comparison Optimization Methods.