One-class classifier

One-class classifier is a type of machine learning model that focuses on learning the feature representation of a single category of data to identify and distinguish new samples belonging to this category from outliers. Its primary objective is to establish an effective boundary that describes the target category, thereby enabling the detection of anomalies. In scenarios where there is a lack of labeled data or insufficient labeled data, one-class classifiers have significant application value and are widely used in anomaly detection, novel class discovery, and other fields. By learning the distribution of normal data, this model can play a crucial role in applications such as financial fraud detection, cybersecurity, and medical diagnosis.