Quantifying the Knowledge in GNNs for Reliable Distillation into MLPs
This work addresses the challenge of efficient inference in graph-based tasks by enhancing distillation reliability, representing an incremental improvement over existing methods.
The paper tackles the problem of reliably distilling knowledge from Graph Neural Networks (GNNs) into Multi-Layer Perceptrons (MLPs) by quantifying node reliability based on information entropy invariance to noise, proposing a method that improves MLPs by 12.62% and outperforms teacher GNNs by 2.16% on average across datasets.
To bridge the gaps between topology-aware Graph Neural Networks (GNNs) and inference-efficient Multi-Layer Perceptron (MLPs), GLNN proposes to distill knowledge from a well-trained teacher GNN into a student MLP. Despite their great progress, comparatively little work has been done to explore the reliability of different knowledge points (nodes) in GNNs, especially their roles played during distillation. In this paper, we first quantify the knowledge reliability in GNN by measuring the invariance of their information entropy to noise perturbations, from which we observe that different knowledge points (1) show different distillation speeds (temporally); (2) are differentially distributed in the graph (spatially). To achieve reliable distillation, we propose an effective approach, namely Knowledge-inspired Reliable Distillation (KRD), that models the probability of each node being an informative and reliable knowledge point, based on which we sample a set of additional reliable knowledge points as supervision for training student MLPs. Extensive experiments show that KRD improves over the vanilla MLPs by 12.62% and outperforms its corresponding teacher GNNs by 2.16% averaged over 7 datasets and 3 GNN architectures.