Skip to main content

Table 2 Classification results achieved on different feature subsets. The optimal values of σ and λ are also given.

From: A machine learning approach for the identification of odorant binding proteins from sequence-derived properties

Features σ λ LOOA BLOOA TP rates TN rates MCC
1463 2.614e-5 1e-009 0.977 0.965 0.945 0.984 0.922
450 4.714e-5 1e-009 0.975 0.963 0.945 0.981 0.915
250 6.325e-5 1e-008 0.970 0.961 0.948 0.975 0.901
100 1e-4 1e-008 0.970 0.962 0.950 0.975 0.903
50 1.414e-4 1e-008 0.967 0.958 0.945 0.971 0.891
  1. LOOA – Leave-one-out accuracy (LOOA); BLOOA – Balanced LOOA MCC – Matthews Correlation Coefficient; σ – Kernel-parameter λ – Regularization parameter; TN – True negative; TP-True positive