Detection and segmentation of changes of dynamics in biological time series

Abstract

A method to identify switching dynamics in time series, based on annealed competition of experts algorithm (ACE), has been developed by J. Kohlmorgen, et al (2000). Incorrect selection of embedding dimension and time delay of the signal significantly affect the performance of the ACE method, however. We utilize systematic approaches based on mutual information and false nearest neighbor to determine appropriate embedding dimension and time delay. Moreover, we obtained further improvements to the original ACE method by incorporating a phase space closeness measure during the training procedure as well as deterministic annealing approach. Using these ameliorated implementations, we have enhanced the performance of the ACE algorithm in determining the location of the switching of dynamic modes in time series. The application of the improved ACE method to RR interval data obtained from rats during control and administration of double autonomic blockade conditions indicate that the improved ACE algorithm is able to segment dynamic mode changes with pinpoint accuracy and that its performance is far superior to the original ACE algorithm.