Skip to main content

Table 6 Optimization of tuning parameter λ for 1-norm SVM

From: Implementation of multiple-instance learning in drug activity prediction

Data set

Cross-validationa

λ

n b

I

0.693

0.001

223

II

0.880

2.000

80

III

0.912

0.016

77

IV

0.598

0.125

89

  1. a The median classification accuracy for 5 replications of 5-fold cross-validation; b the number of important pharmacophore bits.