LOBSTUR: A Local Bootstrap Framework for Tuning Unsupervised Representations in Graph Neural Networks

2025/05/20 by So Won Jeong, Jeong, So Won, Claire Donnat +1
Computer Science · #Advanced Graph Neural Networks #FOS: Computer and information sciences #Graph Theory and Algorithms #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Neural Networks and Applications

paper · pdf · doi:10.48550/arxiv.2505.14867

openalex publication_date 2025/05/20 · openalex created_date 2025/10/19 · openalex updated_date 2026/07/28

Abstract

Graph Neural Networks (GNNs) are increasingly used in conjunction with unsupervised learning techniques to learn powerful node representations, but their deployment is hindered by their high sensitivity to hyperparameter tuning and the absence of established methodologies for selecting the optimal models. To address these challenges, we propose LOBSTUR-GNN (\bf Local \bf Boot\bf strap for \bf Tuning \bf Unsupervised \bf Representations in GNNs) i), a novel framework designed to adapt bootstrapping techniques for unsupervised graph representation learning. LOBSTUR-GNN tackles two main challenges: (a) adapting the bootstrap edge and feature resampling process to account for local graph dependencies in creating alternative versions of the same graph, and (b) establishing robust metrics for evaluating learned representations without ground-truth labels. Using locally bootstrapped resampling and leveraging Canonical Correlation Analysis (CCA) to assess embedding consistency, LOBSTUR provides a principled approach for hyperparameter tuning in unsupervised GNNs. We validate the effectiveness and efficiency of our proposed method through extensive experiments on established academic datasets, showing an 65.9% improvement in the classification accuracy compared to an uninformed selection of hyperparameters. Finally, we deploy our framework on a real-world application, thereby demonstrating its validity and practical utility in various settings. \footnoteThe code is available at \hrefhttps://github.com/sowonjeong/lobstur-graph-bootstrapgithub.com/sowonjeong/lobstur-graph-bootstrap.

Citations

• Graph fission and cross-validation
• Graph-Neural-Network-Based Unsupervised Learning of the Temporal Similarity of Structural Features Observed in Molecular Dynamics Simulations
• Unsupervised Embedding Quality Evaluation
• Uncertainty Quantification over Graph with Conformalized Graph Neural Networks
• Data thinning for convolution-closed distributions
• Graph Neural Networks for Microbial Genome Recovery
• From Canonical Correlation Analysis to Self-supervised Graph Neural Networks
• On Feature Decorrelation in Self-Supervised Learning
• Cross-Validation: What Does It Estimate and How Well Does It Do It?
• Large-Scale Representation Learning on Graphs via Bootstrapping
• Graph Contrastive Learning with Augmentations
• Deep Graph Contrastive Representation Learning
• A Simple Framework for Contrastive Learning of Visual Representations
• Bootstrapping Networks with Latent Space Structure
• Maximum Likelihood Estimation of Sparse Networks with Missing Observations
• Network estimation via graphon with node features
• UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction
• Estimating the number of clusters using cross-validation
• Network cross-validation by edge sampling
• Variational Graph Auto-Encoders
• Network Cross-Validation for Determining the Number of Communities in Network Data
• Cross-Validation for Unsupervised Learning
• Modeling homophily and stochastic equivalence in symmetric relational data
• Cluster Validation by Prediction Strength
• The Stationary Bootstrap

Related

• Self-supervised Graph Representation Learning via Bootstrapping
• Large-scale graph representation learning with very deep GNNs and self-supervision
• Haste Makes Waste: A Simple Approach for Scaling Graph Neural Networks
• Diffusing to the Top: Boost Graph Neural Networks with Minimal Hyperparameter Tuning
• Combining Label Propagation and Simple Models Out-performs Graph Neural Networks