Bayesian Inference of Diffusion Networks with Unknown Infection Times

TL;DR

This paper introduces a Bayesian framework for inferring hidden parameters such as infection times, relationships, and connection strengths in diffusion networks, validated through simulations on synthetic and real data.

Contribution

It presents a novel Bayesian inference method to jointly estimate multiple hidden variables in diffusion networks, addressing a complex real-world problem.

Findings

Effective inference of hidden parameters demonstrated on synthetic data

Successful application to real-world diffusion processes

Framework shows promising accuracy and efficiency

Abstract

The analysis of diffusion processes in real-world propagation scenarios often involves estimating variables that are not directly observed. These hidden variables include parental relationships, the strengths of connections between nodes, and the moments of time that infection occurs. In this paper, we propose a framework in which all three sets of parameters are assumed to be hidden and we develop a Bayesian approach to infer them. After justifying the model assumptions, we evaluate the performance efficiency of our proposed approach through numerical simulations on synthetic datasets and real-world diffusion processes.