Machine self-learning applications in security systems

Abstract

A set of conflict resolution methods is investigated with the purpose to construct knowledge refinement and self-learning tools applicable in a wide range of security systems (SS). The introduction of ontologies simplifies the detection process and lowers the complexity of the machine learning procedures. Different conflict resolution ways lead to particular autonomous/intelligent applications in different types of SS. The proposed self-learning methods are combinable with other web/data mining, anomaly detection, statistical methods, and show new ways in the development of collective evolutionary systems.