One-class tasks

One-class algorithms try to identify outliers and values of a specific class among all the available values, learning from a training set containing the objects of that class, known as positive class or normal samples.

The main aim is to identify outliers in a set of data.

When a one-class algorithm is applied to new data, it imputes a score or an anomaly probability to each sample.

Those which have higher score or lower probability are considered outliers.

• One-class Logic Learning Machine
  • The Options tab
  • The Monitor tab
  • The Results tab
  • Example