Characterization and analysis of physiological time series and biological complex networks

Abstract

Nowadays, there are a large number of time series analysis techniques which allow researchers to summarize the characteristics of a time series into compact metrics, which can then be used to understand the its dynamics or predict how the system will evolve with time. Recently, Campanharo et al. proposed a time series technique based on the complex network theory. More specifically, it is a map from a time series to a complex network which allows to use several networks metrics to characterize the dynamic of a time series. Furthermore, this map has a natural inverse operation, making it possible to map a complex network into a time series. Thus, the topology and dynamic of a given complex network can be explored by analyzing the statistical properties of the corresponding network. In this project we propose an investigation of the map proposed by Campanharo et al. First, using the complex network theory we will distinguish different dynamic regimes of EEG time series from healthy and unhealthy subjects. Moreover, using the time series analysis we will investigate the self-similarity in complex networks. More specifically, the topology and dynamic of several synthetic and biological complex networks, such as metabolic and protein-protein interaction networks, will be investigated using the time series analysis. (AU)