Abstract: This work deals in the usage of fractal analysis to determine the trend of dynamic characteristics of chaotic series based on R/S-analysis. We propose an algorithm using the method of dynamic roles with large dimensions, which can be specified Hurst at every step of aggregation without storing values of data stream based on neural network technology.

Keywords: Complex System of Structured Type (CSST), chaotic dynamic series, fractal analysis, Hausdorff rate, Hurst rate, algorithm R/S -analysis, persistence, antipersistence, prognosis, neural network.

ACM Classification Keywords: G.1.2 Approximation, F.1.1 Models of Computation (F.4.1)

Link:

ALGORITHMIZATION PROCESS FOR FRACTAL ANALYSIS IN THE CHAOTIC DYNAMICS OF COMPLEX SYSTEMS STRUCTURED TYPE

Igor Skiter, Еlena Trunova

http://www.foibg.com/ijita/vol22/ijita22-03-p06.pdf