QUALITY IMPROVEMENT OF INFORMATION SECURITY EVENTS IDENTIFICATION THROUGH INPUT DATA SPLITTING
I. S. Lebedev, M. E. Sukhoparov, D. D. Tikhonov Saint Petersburg Federal Research Center of Russian Science Academy Russian State Hydrometeorological University
Annotation: The processing of information sequences using segmentation of input data is proposed, aimed at improving the quality of detection of destructive influences using machine learning models. The basis of the proposed solution is the division of data into segments with different properties of the objects of observation. A method using a multi-level data processing architecture is described, where learning processes are implemented at various levels, the analysis of the achieved values of quality indicators and the assignment of the best models for quality indicators to individual data segments. The proposed method makes it possible to improve the quality indicators for detecting destructive information influences by segmenting and assigning models that have the best performance in individual segments
Keywords: information security, machine learning, data set, data sampling, data segmentation, processing models
Pages 31–43