Nonparametric Identification of Kronecker Networks

TL;DR

This paper introduces a kernel-based PEM method for estimating dynamic networks with Kronecker-structured topology, enabling better understanding of complex network organization through nonparametric identification.

Contribution

The paper presents a novel kernel-based PEM approach specifically designed for nonparametric identification of Kronecker-structured dynamic networks.

Findings

Method effectively estimates Kronecker networks

Numerical examples demonstrate high accuracy

Approach enhances understanding of complex network organization

Abstract

We address the problem to estimate a dynamic network whose edges describe Granger causality relations and whose topology has a Kronecker structure. Such a structure arises in many real networks and allows to understand the organization of complex networks. We proposed a kernel-based PEM method to learn such networks. Numerical examples show the effectiveness of the proposed method.