Supplemental Material: Probability Models for Open Set Recognition

  title={Supplemental Material: Probability Models for Open Set Recognition},
  • Published 2014
A Results for Multi-class Open Set Recognition Accuracy The main article presents F-measure plots, but reviewers may be interested in accuracy plots for comparison. Accuracy is a common choice for evaluating binary decision classifiers. However, it is not always the best choice because it tends to underemphasize the distinction between correct positive and negative classification. Precisely defined, accuracy refers to the correctly classified samples (true positives TP and true negatives TN… CONTINUE READING