The Probabilistic Backbone of Data-Driven Complex Networks: An example in Climate

TL;DR

This paper advocates using Bayesian Networks to construct data-driven complex networks, like climate networks, because they reduce redundancy and capture essential probabilistic relationships, providing a more meaningful system representation.

Contribution

The paper introduces Bayesian Networks as a probabilistic backbone for complex networks, improving the extraction of physical features from climate data.

Findings

Bayesian Networks produce sparser, more meaningful climate networks.

BNs effectively reduce redundant information compared to correlation networks.

Application to climate data demonstrates the method's utility.

Abstract

Correlation Networks (CNs) inherently suffer from redundant information in their network topology. Bayesian Networks (BNs), on the other hand, include only non-redundant information (from a probabilistic perspective) resulting in a sparse topology from which generalizable physical features can be extracted. We advocate the use of BNs to construct data-driven complex networks as they can be regarded as the probabilistic backbone of the underlying complex system. Results are illustrated at the hand of a global climate dataset.