Graph Sparsification for Enhanced Conformal Prediction in Graph Neural Networks

2024/10/28 by Yuntian He, Pranav Maneriker, He, Yuntian +7
Computer Science · #Advanced Graph Neural Networks #Neural Networks and Applications #Graph Theory and Algorithms

paper · pdf · doi:10.48550/arxiv.2410.21618

Abstract

Conformal Prediction is a robust framework that ensures reliable coverage across machine learning tasks. Although recent studies have applied conformal prediction to graph neural networks, they have largely emphasized post-hoc prediction set generation. Improving conformal prediction during the training stage remains unaddressed. In this work, we tackle this challenge from a denoising perspective by introducing SparGCP, which incorporates graph sparsification and a conformal prediction-specific objective into GNN training. SparGCP employs a parameterized graph sparsification module to filter out task-irrelevant edges, thereby improving conformal prediction efficiency. Extensive experiments on real-world graph datasets demonstrate that SparGCP outperforms existing methods, reducing prediction set sizes by an average of 32% and scaling seamlessly to large networks on commodity GPUs.

Related

• Improving the interpretability of GNN predictions through conformal-based graph sparsification
• RoCP-GNN: Robust Conformal Prediction for Graph Neural Networks in Node-Classification
• Triple Sparsification of Graph Convolutional Networks without Sacrificing the Accuracy
• Learnt Sparsification for Interpretable Graph Neural Networks
• Residual Reweighted Conformal Prediction for Graph Neural Networks