Connectivity Analysis for Multivariate Time Series: Correlation vs. Causality

Abstract

The study of the interdependence relationships of the variables of an examined system is of great importance and remains a challenging task. There are two distinct cases of interdependence. In the first case, the variables evolve in synchrony, connections are undirected and the connectivity is examined based on symmetric measures, such as correlation. In the second case, a variable drives another one and they are connected with a causal relationship. Therefore, directed connections entail the determination of the interrelationships based on causality measures. The main open question that arises is the following: can symmetric correlation measures or directional causality measures be applied to infer the connectivity network of an examined system? Using simulations, we demonstrate the performance of different connectivity measures in case of contemporaneous or/and temporal dependencies. Results suggest the sensitivity of correlation measures when temporal dependencies exist in the data. On the other hand, causality measures do not spuriously indicate causal effects when data present only contemporaneous dependencies. Finally, the necessity of introducing effective instantaneous causality measures is highlighted since they are able to handle both contemporaneous and causal effects at the same time. Results based on instantaneous causality measures are promising; however, further investigation is required in order to achieve an overall satisfactory performance.

The Causality and Uncertainty of the COVID-19 Pandemic to Bursa Malaysia Financial Services Index’s Constituents

“…From a high-level view, the difference in the outcomes between G-causality and Schreiber transfer entropy are possibly due to the non-linear nature of transfer entropy and its ability to detect outliers [ 109 , 110 ]. If the financial market is viewed as a complex system, non-linearity [ 109 , 111 ] and outliers in view of power laws [ 112 ] are to be expected.…”

Transfer Entropy and Granger Causality in Real-world Telecommunication Networks

“…With transfer entropy primarily applicable to bivariate data of only 1s and 0s, real world telco network typically consists of multivariate data [3]. Compensated Transfer Entropy (cTE) is observed to be a reliable method to estimate information transfer in physiological multi-variate time series data as described in [12].…”

Ordinal methods for a characterization of evolving functional brain networks

“…via probing), one usually resorts to linear and non-linear time-series-analysis techniques to quantify interaction properties from pairs of time series of appropriate system observables. These techniques originate from diverse fields such as statistics, synchronization theory, non-linear dynamics, information theory, statistical physics, and from the theory of stochastic processes [44][45][46][47][48][49][50][51][52][53][54][55][56] , given that interactions can manifest themselves in various aspects of the dynamics. While the majority of studies on (evolving) functional networks is based on binary (an edge exists or not) or weighted networks (the weight of an edge is given by the strength of interaction), further improvements can be expected by considering weighted and directed networks, thereby including knowledge about coupling functions that contain detailed information about the functional mechanisms underlying an interaction and that prescribe the physical rule specifying how an interactions occurs 57 .…”