Finding Hidden Communities in Complex Networks from Chaotic Time Series

Abstract

Recent works show that complex network theory may be another powerful tool in time series analysis. In this paper, we construct complex networks from the chaotic time series with Maximal Information Coefficient (MIC). Each vector point in the reconstructed phase space is represented by a single vertex and edge determined by MIC. By using the Chua’s circuit system, we illustrate the potential of these complex network measures for the detection of the topology structure of the network. Comparing with the linear relationship measure, we find that the topology structure of the community with MIC reveals the hidden or implied correlation of the network.