IMPROVING THE QUALITY OF INFORMATION SECURITY STATE IDENTIFICATION BASED ON SAMPLE SEGMENTATION
I. S. Lebedev, M. E. Sukhoparov St. Petersburg Federal Research Centre of the Russian Academy of Sciences, Russian State Hydrometeorological University
Annotation: Improving the quality indicators of information security state identification of individual segments of cyber-physical systems is associated with the processing of large information arrays. We propose a method for improving quality indicators when solving problems of information security state identification. Its implementation is based on the formation of individual segments of the sample. The analysis of properties of these segments makes it possible to select and assign algorithms having the best quality indicators on the current segment. The segmentation of the data sample is considered. Experimental values of the quality index for the proposed method for different classifiers on individual segments and the whole sample are given on the example of real dataset data.
Keywords: information security, machine learning, data set, data segmentation, data sampling.