Skip to content

PreGS: A Parameter-Transfer-Based Multi-Expert Graph Neural Network for Node Classification

Sep 2026 · 0 citations · 33 references
Computer Science

TL;DR

This paper proposes PreGS, a parameter-transfer-based multi-expert graph neural network framework, and develops PreGSv2, which introduces source-level weighting and a structural gating mechanism for adaptive multi-source feature integration.

Abstract

Graph neural networks have achieved strong performance in node classification by aggregating information from graph neighborhoods. However, a single aggregation mechanism may be insufficient to capture diverse structural patterns across graph datasets. Moreover, independently training multiple structural branches can introduce substantial overhead without necessarily producing stable node representations. To address these issues, this paper proposes PreGS, a parameter-transfer-based multi-expert graph neural network framework. PreGS first pretrains a multi-head graph attention network (GAT) and transfers the linear transformation weights of its first-layer attention heads to multiple GraphSAGE experts. The transferred experts are frozen and used as complementary structural branches. The fused raw node features, GAT head representations, and GraphSAGE expert representations are fed into a multilayer perceptron (MLP), whose output is further fused with the pretrained GAT logits. Based on PreGS, we further develop PreGSv2, which introduces source-level weighting and a structural gating mechanism for adaptive multi-source feature integration. Experiments on eight public graph datasets show that PreGS and PreGSv2 achieve competitive performance against representative graph neural network baselines. Ablation, parameter-transfer, sensitivity, aggregator, visualization, and training-time analyses further validate the effectiveness and stability of the proposed framework. The code and datasets are available at https://github.com/LH-Czc/PreGS.

View source

Similar papers

Conference Open access Sep 2026

ADC-GNN: Adaptive Dual-level Collaborative Graph Neural Networks for Graph Classification

This work proposes a novel Adaptive Dual-level Collaborative GNN associated with an adaptive dual-level collaborative mechanism, and shows that the ADC-GNN can inject the learned high-level information back into the node level, forming a closed-loop, bidirectional optimization process.

Wan Tang, Lu Bai, Li-Xin Cui et al. · 0 citations
Preprint Aug 2026

Why Does Graph Learning Fail to Fully Benefit from a Text Teacher?

A multimodal model that combines two complementary ideas: a self-supervised method that enables a GNN encoder pretrained on one dataset to operate directly on another dataset with a different node-feature dimensionality, without rebuilding the model or realigning the data is investigated.

Fumiaki Kimino, Ryoma Sato Sokendai, National Institute of Informatics · 1 citation

GEM-Distill: Graph Ensemble Multi-Level Distillation for MLP-Based Graph Classification

Graph Neural Networks (GNNs) achieve strong performance on graph learning tasks, but their message-passing computation hinders deployment in resource-constrained settings. GNN-to-MLP (G2M) knowledge distillation improves inference efficiency by transferring GNN knowledge to lightweight MLP students; however, existing m...

Ming-Hao Qin, Dong-Hai Guan, Weiwei Yuan et al. · 0 citations

Expert Systems With Applications

A Multi-Relation Enhanced Graph Disentangled Network (MREGDN) is proposed, which can effectively disentangle the latent factors that affect graph topology, while enhancing the interpretability and low-redundancy of the learned node representations for classification.

Xin-Xin Liu, Wei-Ren Yu · 0 citations
Open access Sep 2026

GTAP: Graph Topology-Aware Pre-Training for Graph Classification

GTAP (Graph Topology-Aware Pre-training), a self-supervised initialization framework for within-dataset graph classification, improves over a matched GCN trained from scratch and achieves competitive accuracy against published baselines on most datasets with available results.

Chao-Chao Hu, Zhao-Hui Zhang · 0 citations
Preprint Aug 2026

Two-level domain-decomposition AdaGrad method for scalable training of graph neural networks

The proposed DD-AG2m alternates between AG2m optimization on the original (global) graph and AG2m optimization on the partitioned graphs, and introduces a two-level variant that performs global optimization steps on a coarse graph obtained by randomly subsampling nodes within each subdomain.

Laurynas Varnas, Julien Herrmann, Alexander Heinlein et al. · 0 citations

Related blog posts

MIT News · Artificial Intelligence Nov 17, 2026

Learning the language of machine learning

By studying the computational foundations of language learning, Associate Professor Jacob Andreas builds intelligent systems that can communicate effectively with humans.

Microsoft Research Blog Nov 16, 2026

TokenEconomizer: Learning from Experience to Make AI Coding Agents More Efficient

AI coding agents are increasingly help developers fix bugs and build features largely on their own, but every interaction with their models consumes tokens. In each session, an agent sends instructions, tool descriptions, and session history to its model. Most efforts to reduce this cost focus on individual tasks. Yet developers return to the same […] The post TokenEconomizer: Learning from Experience to Make AI Coding Agents More Efficient appeared first on Microsoft Research.

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.