Ranking Support Vector Machine with Kernel Approximation

Table 4

Performance comparison on OHSUMED data set.

	NDCG@1	NDCG@3	NDCG@5	P@1	P@3	MAP
RankSVM-Struct [9]	0.5515	0.4850	0.4729	0.6338	0.5898	0.4478
ListNet [19]	0.5326	0.4732	0.4432	0.6524	0.6016	0.4457
AdaRank-MAP [20]	0.5388	0.4682	0.4613	0.6338	0.5895	0.4487
AdaRank-NDCG [20]	0.5330	0.4790	0.4673	0.6719	0.5984	0.4498
RankBoost [18]	0.4632	0.4555	0.4494	0.5576	0.5609	0.4411
RankRLS [24]	0.5490	0.4770	0.4530	0.6440	0.5860	0.4470
RankSVM-Primal [8]	0.5645	0.5004	0.4782	0.6710	0.6112	0.4439
RankNystöm	0.5730	0.4874	0.4780	0.6801	0.5890	0.4473
RankRandomFourier	0.5728	0.4965	0.4804	0.6801	0.5983	0.4472