Summary
International Symposium on Nonlinear Theory and Its Applications
2015
Session Number:B3L-F
Session:
Number:B3L-F-3
Revisiting Surrogate Generation for Cyclic Time Series
Michael Small, Jiahao Su, Ming-Wai Chan,
pp.640-643
Publication Date:2015/12/1
Online ISSN:2188-5079
DOI:10.34385/proc.47.B3L-F-3
PDF download (949KB)
Summary:
The cycle shuffled surrogate algorithm provides a straightforward method to randomise cyclic time series. This randomisation can then be employed as a form of Monte-Carlo hypothesis testing --- do the randomised realisations differ, statistically, from the original? If they do, then one may conclude (with some additional caveats) that the original data included deterministic inter-cycle dynamics: deterministic chaos, for example. In this communication we will re-examine this algorithm, point to several technical issues that may arise and discuss suitable palliatives.