Abstract: Anomaly detection needs to learn one-class classifiers from normal instances in observation or feature spaces. In the Neyman–Pearson criterion, the design of one-class classifiers boils down ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results