Summary
International Symposium on Nonlinear Theory and its Applications
2009
Session Number:A1L-D
Session:
Number:A1L-D4
A Method for Transforming Complex Networks to Time Series Using Classical Multidimensional Scaling
Yuta Haraguchi, Yutaka Shimada, Tohru Ikeguchi,
pp.-
Publication Date:2009/10/18
Online ISSN:2188-5079
Summary:
Novel methods have been proposed to analyze nonlinear dynamical systems using the complex network theory. In these methods, time series are transformed to complex networks and the networks are analyzed by the methods based on the complex network theory. In this paper, we propose an opposite direction of these methods: we transform complex networks to time series by using classical multidimensional scaling which is one of the multivariate analysis methods. Then, we analyze transformed time series using the nonlinear time series analysis methods. In numerical simulations, using two mathematical models which can generate regular, random, and small-world networks, we examine the characteristics of the transformed time series from these two network models. We show the structural difference between two small-world networks generated from the two different models. This result indicate that our method can reveal a hidden structural property in the complex networks.