Data Anomaly Detection through Semisupervised Learning Aided by Customised Data Augmentation Techniques

Figure 7

DA samples for “outlier” and “drift.”