Complex algorithm for detecting anomalous features in natural data

A complex automated algorithm for analyzing natural data and detecting anomalous features is proposed. The algorithm includes an algorithm for determining the information components of a signal and an algorithm for adaptive wavelet filtering of a signal. The algorithm for determining the information components of a signal suppresses correlated noise and determines the information components of a signal. The algorithm for adaptive wavelet filtering of a signal detects anomalous features and estimates their intensity. The algorithms are based on the rules developed by the authors. Based on the rules, the parameters of the threshold function are estimated and the best approximating wavelet is determined. The article describes the operations of the complex algorithm and presents a block diagram of its implementation. Also presented are the results of applying a complex algorithm using data from secondary cosmic rays and model data constructed in their likeness. The results confirmed the effectiveness of the developed rules and the proposed complex algorithm.