Figure 3.

Comparison of PSI-BLAST and RankProp variants. Each figures plots the number of test set queries for which a given method achieves a ROC_n score threshold. Figures (A), (B) and (C) use ROC₅₀, ROC₁₀ and ROC₁ scores, respectively. Corresponding mean ROC_n scores are given in Table 1. The variants of RankProp are described in the text. Previous work [5] used RankProp with σ = 100.