Gradients of O-information: low-order descriptors of high-order dependencies

Scagliarini, T, Nuzzi, D, Antonacci, Y, Faes, L, Rosas, F E, Marinazzo, D and Stramaglia, S (2023) Gradients of O-information: low-order descriptors of high-order dependencies. Physical Review Research, 5. a013025 1-8. ISSN 2643-1564

[img] PDF - Published Version
Available under License Creative Commons Attribution.

Download (882kB)


O-information is an information-theoretic metric that captures the overall balance between redundant and synergistic information shared by groups of three or more variables. To complement the global assessment provided by this metric, here we propose the gradients of the O-information as low-order descriptors that can characterize how high-order effects are localized across a system of interest. We illustrate the capabilities of the proposed framework by revealing the role of specific spins in Ising models with frustration, in Ising models with three-spin interactions, and in a linear vectorial autoregressive process. We also provide an example of practical data analysis on U.S. macroeconomic data. Our theoretical and empirical analyses demonstrate the potential of these gradients to highlight the contribution of variables in forming high-order informational circuits.

Item Type: Article
Schools and Departments: School of Engineering and Informatics > Informatics
SWORD Depositor: Mx Elements Account
Depositing User: Mx Elements Account
Date Deposited: 28 Feb 2023 09:56
Last Modified: 28 Feb 2023 10:00

View download statistics for this item

📧 Request an update