The Cumulative Sum Algorithm as a Means for Detecting of the Time Series Disorder with a One-dimensional Exponential Distribution

The issue of detection in real timescale a sudden abrupt change in the properties (disorder) of a time series with a one-dimensional exponential probability distribution function is considered. It is proposed to use the CUSUM detection algorithm to solve this problem. Using the simulation method, complete information was obtained that ensures the synthesis of the CUSUM detection algorithm with user-defined characteristics. The possibilities of using the information obtained during laboratory work on the course "Analysis of Stochastic processes" in the framework of the educational process at the Department of Management and Intelligent Technologies of the National Research University MPEI.