Detecting structural changes on time series

  1. María Muñoz-Guillermo
  2. G. García-Clemente
  3. J. Cánovas
Actas:
10th Chaotic Modeling and Simulation International Conference: Barcelona, del 30 de mayo al 2 de junio de 2017

Año de publicación: 2017

Tipo: Aportación congreso

GOOGLE SCHOLAR

Resumen

Entropy has been revealed as a measure of the complexity to detect thestructural changes applied to describe the dynamics. We considerdifferent entropy-like measures to detect structural changes in a timeseries. We analyze the results of different of these measures topreviously known generated time series to compare the normalizedresults and determine what measure fits better. An application to a realseismic data series is shown.