Learning Sparse Graphs via Majorization-Minimization for Smooth Node Signals

TL;DR

This paper introduces a hyperparameter-free algorithm based on majorization-minimization for efficiently learning sparse graphs from signals that vary smoothly over the nodes, with applications demonstrated on synthetic and brain-network data.

Contribution

It presents a novel MM-based method for sparse graph learning that automatically eliminates inactive variables, improving convergence speed without hyperparameter tuning.

Findings

Faster convergence compared to existing methods.

Effective in synthetic and real-world brain-network data.

No need for hyperparameter tuning.

Abstract

In this letter, we propose an algorithm for learning a sparse weighted graph by estimating its adjacency matrix under the assumption that the observed signals vary smoothly over the nodes of the graph. The proposed algorithm is based on the principle of majorization-minimization (MM), wherein we first obtain a tight surrogate function for the graph learning objective and then solve the resultant surrogate problem which has a simple closed form solution. The proposed algorithm does not require tuning of any hyperparameter and it has the desirable feature of eliminating the inactive variables in the course of the iterations - which can help speeding up the algorithm. The numerical simulations conducted using both synthetic and real world (brain-network) data show that the proposed algorithm converges faster, in terms of the average number of iterations, than several existing methods in the literature.