MS40 ~ Wednesday, May 24, 1995 ~ 10:00 AM

Detection and Classification of Chaotic Signals

Time series analysis methods derived from Nonlinear Dynamics (NLD) have resulted in new signal processing techniques for chaotic or, more generally, nonlinear data. For nearly a decade, applications of NLD have been developed for prediction, noise reduction, and signal separation. More recently, efforts have been directed towards detection and classification of signals, or alternately, of the generating physical system. The speakers in this minisymposium will stress techniques and applications of NLD methods to detect and classify deterministic signal components, or perform system identification, using nonlinear information and discuss methods such as local and global dynamical models, entropies, recurrence plots, and topological classification schemes.

Organizer: James Kadtke, University of California, San Diego

Detection and Classification of Chaotic Signals Using Global Dynamical Models
James Kadtke, Organizer
System Identification in Experimental Data
Stephen Hammel, Naval Surface Warfare Center
Wavelet 40 Hz EEG Coincidences During Associative Conditioning
Gottfried Mayer-Kress, University of Illinois, Urbana
How Many Times Should I Iterate?
Guan-Hsong Hsu, Naval Surface Warfare Center