Segmentation of data when analyzing the state of telecommunication systems

<span lang="EN-US">The identification of abnormal situations in information and telecommunication systems is considered, based on analyze statistical network traffic packages. The method identifying an anomalous situation segmentation data sample proposed. aimed at using classifying algorithms that have the best quality indicators individual segments. proposed will be useful for monitoring security systems. registers factors affect change properties targeted variables. Impact detection allows you to generate samples, depending current expected situations. On example NSL-KDD dataset, there was a division many into subset, taking account influence range values. processing shown point function time series. With its use, by final number non-intersecting measurable subsets has been made. results Accuracy, Precision, F-Measure, Recall various classifiers are shown. increase classification continuously changing operating conditions systems.</span>