Deep Iterative and Adaptive Learning for Graph Neural Networks

TL;DR

This paper introduces DIAL-GNN, an end-to-end framework that jointly learns graph structures and embeddings through iterative, adaptive methods, improving performance and efficiency in graph-based learning tasks.

Contribution

The paper presents a novel iterative approach for adaptive graph structure learning integrated with GNNs, outperforming existing methods in accuracy and computational efficiency.

Findings

DIAL-GNN outperforms state-of-the-art baselines in various tasks.

The method effectively balances smoothness, connectivity, and sparsity.

It handles both transductive and inductive learning scenarios.

Abstract

In this paper, we propose an end-to-end graph learning framework, namely Deep Iterative and Adaptive Learning for Graph Neural Networks (DIAL-GNN), for jointly learning the graph structure and graph embeddings simultaneously. We first cast the graph structure learning problem as a similarity metric learning problem and leverage an adapted graph regularization for controlling smoothness, connectivity and sparsity of the generated graph. We further propose a novel iterative method for searching for a hidden graph structure that augments the initial graph structure. Our iterative method dynamically stops when the learned graph structure approaches close enough to the optimal graph. Our extensive experiments demonstrate that the proposed DIAL-GNN model can consistently outperform or match state-of-the-art baselines in terms of both downstream task performance and computational time. The proposed approach can cope with both transductive learning and inductive learning.