Partial Information Decomposition of Boolean Functions: a Fourier Analysis perspective

Abstract: Partial information decomposition (PID) partitions the information that a set of sources has about a target variable into synergistic, unique, and redundant contributions. This information-theoretic tool has recently attracted attention due to its potential to characterize the information processing in multivariate systems. However, the PID framework still lacks a solid and intuitive interpretation of its information components. In the aim to improve the understanding of PID components, we focus here on Boolean gates, a much-studied type of source-target mechanisms. Boolean gates have been extensively characterised via Fourier analysis which coefficients have been related to interesting properties of the functions defining the gates. In this paper, we establish for Boolean gates mechanisms a relation between their PID components and Fourier coefficients.