A number of practical tasks in engineering, finance, and medicine can be formulated as a problem of imbalanced classification and a problem of anomaly detection, which we can solve it by describing the normal state of the object using, e.g., a one-class classification model. In the presentation, we provide a brief overview of models, methods, and new approaches to imbalanced classification and anomaly detection. Particular attention is paid to the practical use of predictive maintenance methods for various engineering systems: monitoring of the technical condition of a power unit, forecasting parameters of the roadway on basis of data from road weather stations, etc.
