INVESTIGADORES
LUCINI Maria magdalena
artículos
Título:
The asymptotic distribution of the permutation entropy
Autor/es:
REY, A. A.; FRERY, A. C.; GAMBINI, J.; LUCINI, M. M.
Revista:
CHAOS AN INTERDISCIPLINARY JR OF NONLINEAR SCIENCE
Editorial:
AMER INST PHYSICS
Referencias:
Lugar: New York; Año: 2023 vol. 33
ISSN:
1054-1500
Resumen:
Ordinal Patterns serve as a robust symbolic transformation technique, enablingthe unveiling of latent dynamics within time series data. This methodology involves constructing histograms of patterns, followed by the calculation of both entropy and statistical complexity an avenue yet to be fully understood in termsof its statistical properties. While asymptotic results can be derived by assuminga Multinomial distribution for histogram proportions, the challenge emerges fromthe non-independence present in the sequence of ordinal patterns. Consequently,the direct application of the Multinomial assumption is questionable. This studyfocuses on the computation of the asymptotic distribution of permutation entropy, considering the inherent patterns correlation structure. Furthermore, theresearch delves into a comparative analysis, pitting this distribution against theentropy derived from a Multinomial law. We present simulation algorithms forsampling time series with prescribed histograms of patterns and transition probabilities between them. Through this analysis, we better understand the intricaciesof ordinal patterns and their statistical attributes.
