A clustering method has been developed to group signals that display similar dynamic behavior. The procedure involves using the method of time delay embedding to construct a trajectory in state space from a time series. Certain features that characterize the geometry of the trajectory have been defined. These features were subjected to a series of statistical tests to determine their usefulness in a hierarchical clustering analysis. The latter is aimed at finding groups of similar trajectories. The trajectory‐based clustering algorithm has been applied to simulated data, which included both stochastic data generated by a linear AR model, and nonlinear data generated by a Duffing oscillator. The results show that the algorithm works reliably in both cases.

1.
A.
Wolf
,
J. B.
Swift
,
H. L.
Swinney
, and
J. A.
Vastano
,
Physica D
16
,
285
(
1985
).
Google Scholar
Crossref
Search ADS
 
2.
J. D.
Farmer
,
A.
Ott
, and
J. A.
Yorke
,
Physica D
7
,
153
(
1983
).
Google Scholar
Crossref
Search ADS
 
3.
P.
Grassberger
 and
I.
Procaccia
,
Physica D
9
,
189
(
1983
).
Google Scholar
Crossref
Search ADS
 
4.
P.
Grassberger
 and
I.
Procaccia
,
Phys. Rev. Lett.
50
,
346
(
1983
).
Google Scholar
Crossref
Search ADS
 
5.
J.
Theiler
,
Phys. Rev. A
34
,
2427
(
1986
).
Google Scholar
Crossref
Search ADS
 
6.
Y. Gong and J. P. Haton, “Text-independent speaker recognition by trajectory space comparison,” International Conference on Acoustics, Speech and Signal Processing 90, Vol. 1, pp. 285–288, Albuquerque, 1990.
7.
F. C. Richards, T. C. Bourgeois, and N. H. Packard, Internal Report, Center for Complex Systems Research, University of Illinois, 13 pages, 1990 (unpublished).
8.
N. H.
Packard
,
J. P.
Crutchfield
,
J. D.
Farmer
, and
R. S.
Shaw
,
Phys. Rev. Lett.
45
,
712
(
1980
).
Google Scholar
Crossref
Search ADS
 
9.
A.
Fraser
 and
H.
Swinney
,
Phys. Rev. Lett. A
33
,
1134
(
1986
).
Google Scholar
Crossref
Search ADS
 
10.
M. B.
Kennel
,
R.
Brown
, and
H. D. I.
Abarbanel
,
Phys. Rev. A
45
,
3403
(
1992
).
Google Scholar
Crossref
Search ADS
PubMed
 
11.
Because a trajectory may pass through the cell without any state space vectors pointing to that cell, or because several state space vectors, on the same trace of the trajectory, may reside in the same cell, Gk(j) and Mk(j) may differ.
12.
D. T.
Kaplan
 and
L.
Glass
,
Phys. Rev. Lett.
68
,
427
(
1992
).
Google Scholar
Crossref
Search ADS
PubMed
 
13.
R. O. Duda and P. E. Hart, Pattern Classification and Scene Analysis (Wiley, New York, 1973).
14.
J. H.
Ward
,
J. Am. Statist. Assoc.
58
,
236
(
1963
).
Google Scholar
Crossref
Search ADS
 
15.
W. J. Dixon and M. B. Brown, BMDP-79, Biomedical Computer Programs P-Series (University of California, Berkeley, 1979).
16.
M. James, Classification Algorithms (Wiley, New York, 1985).
17.
G. Duffing, Erzwungene Schwingungen bei Veränderlicher Eigenfrequenz (Vieweg, Braunschweig, 1918).
18.
Y. Ueda, “Steady motions exhibited by Duffing’s equation: a picture book of regular and chaotic motions.” in New Approaches to Nonlinear Problems in Dynamics, edited by P. J. Holmes (SIAM, Philadelphia, 1980), pp. 311–322.
19.
H. N. Nyberg, MSc thesis, Department of Electrical Engineering, University of Houston, Houston, TX, 1991.
20.
Determined by way of confidence interval test for the true correlation ρ, given a measured correlation r (Ref. 21, p. 224) (sample size=45) at a confidence level of 0.99.
21.
W. J. Dixon and F. J. Massey, Introduction to Statistical Analysis (McGraw-Hill, New York, 1983).
22.
W. W. Cooley and P. R. Lohnes, Multivariate Data Analysis (Wiley, New York, 1971).
This content is only available via PDF.
© 1993 American Institute of Physics.
1993
American Institute of Physics
You do not currently have access to this content.