It is a one-class classification scenarios or two-classes classification scenarios?

Because this is a Anomaly Detection problem. So when you do the classification, in the training stage, you only use one-class-labeled data? or use both classes of data for anomaly?

21 July 2017 3,886 4 View