Capacity method for analyzing rare events, error estimation due to competition or data loss

Abstract

Purpose: the article briefly describes the idea of a capacity method for analyzing rare events. The article focusses on the drop in accuracy as a result of the data loss. Discussion: error estimation was carried out using computer simulation. The authors describe experimental model, In addition, estimate the error by classical methods, when rare events are represented by time series with zero values. As an estimate of the error, the authors use the average relative error and the mean square relative error. Results: the study shows that the capacity method of rare events analysis has better accuracy compared to the classical method, regardless of how much of the data is lost, 10% 30% or 50%. However, in the case of frequent events, the error of the classical method is almost the same. The authors conclude that for the rare events analysis it is necessary to switch to the capacity method.