Industrial Automatic Control Systems and Controllers Annotation << Back Application of Artifi cial Neural Networks in Intrusion Detection System Development A.G. Mustafayev Intrusion detection systems classify network traffi c into two main categories: normal activity and the actions of an attacker. Currently, intelligent data processing and machine learning play an important role in many areas of activity, not excluding intrusion detection systems. One of the main steps in data mining is the identification of an optimal data set that helps improve the effi ciency, performance and speed of predicting intrusion detection systems. For the experimental analysis, a set of data from the NSL-KDD database was used. The results of the experiments show that the approach proposed in the paper is accurate enough, with a low number of false positives and high sensitivity, requiring less training time than using a complete set of data. Keywords: intrusion detection system; adaptability; classifi cation; artificial neural networks; analysis of network traffic; computer networks. Contacts: E-mail: arslan_mustafaev@mail.ru Pp. 17-26.