The Quality of Ranking during Simulated Pairwise Judgments for Examined Approximation Procedures

Table 4

Mean performance measures of arbitrarily selected APs for 30,000 cases.

Scenario details			Procedure	MRE	Rank	MSRC	Rank	MRR	Rank	Mean rank
Geometric scale	Gamma distribution	FR-PCM	LLSM	0.438438	1	0.682300	2	1.21242	1	1.3 (3)
			REV	0.452614	5	0.668380	5	1.22051	4	4.6 (6)
			LUA	0.447349	2	0.673067	3	1.21792	2	2.3 (3)
			SRDM	0.448759	3	0.671380	4	1.21870	3	3.3 (3)
			SNCS	0.450734	4	0.692453	1	1.24398	5	3.3 (3)
	Uniform distribution	FR-PCM	LLSM	0.288608	1	0.804860	2	1.12813	1	1.3 (3)
			REV	0.302346	4	0.792580	5	1.13530	4	4.3 (3)
			LUA	0.298401	2	0.795767	3	1.13350	2	2.3 (3)
			SRDM	0.299400	3	0.794820	4	1.13400	3	3.3 (3)
			SNCS	0.303463	5	0.808333	1	1.15450	5	3.6 (6)

Note: FR-PCM denotes forced reciprocity applied to PCM during simulations.