Improving the false nearest neighbors method with graphical analysis
T. Aittokallio, M. Gyllenberg, J. Hietarinta, T. Kuusela, T. Multamaki
Abstract
We introduce a graphical presentation for the false nearest neighbors (FNN) method. In the original method only the percentage of false neighbors is computed without regard to the distribution of neighboring points in the time-delay coordinates. With this new presentation it is much easier to distinguish deterministic chaos from noise. The graphical approach also serves as a tool to determine better conditions for detecting low dimensional chaos, and to get a better understanding on the applicability of the FNN method.
Create a lesson
Related papers
Chaotic eigenfunctions in phase space
S. Nonnenmacher, A. Voros
Improved control of delayed measured systems
Jens Christian Claussen, Heinz Georg Schuster
The accurate and comprehensive model of thin fluid flows with inertia on curved substrates
A. J. Roberts, Zhenquan Li
On periodic solutions of a Hamilton-Jacobi equation with periodic forcing
Andrei Sobolevskii
Drifters dispersion in the Adriatic Sea: Lagrangian data and chaotic model
Guglielmo Lacorata, Erik Aurell, Angelo Vulpiani
Generalized multibaker maps for open dissipative systems
Z. Kaufmann, P. Szépfalusy