
Inductive inference model of anomaly and misuse detection
Author(s) -
Paul Helman
Publication year - 1997
Language(s) - English
Resource type - Reports
DOI - 10.2172/427002
Subject(s) - inference , anomaly detection , inductive reasoning , computer science , selection (genetic algorithm) , set (abstract data type) , anomaly (physics) , model selection , data mining , base (topology) , artificial intelligence , machine learning , mathematics , mathematical analysis , physics , programming language , condensed matter physics
Further consequences of the inductive inference model of anomaly and misuse detection are presented. The results apply to the design of both probability models for the inductive inference framework and to the design of W&S rule bases. The issues considered include: the role of misuse models M{sub A}, the selection of relevant sets of attributes and the aggregation of their values, the effect on a rule base of nonmaximal rules, and the partitioning of a set of attributes into a left hand and right hand side