Conditional Uncertainty Quantification for Tensorized Topological Neural Networks

TL;DR

This paper introduces CF-T2NN, a novel method combining tensor decomposition and topological learning to improve uncertainty quantification in graph neural networks, leading to more reliable and interpretable graph classification.

Contribution

The paper presents a new approach, CF-T2NN, for rigorous uncertainty quantification in GNNs that reduces label prediction set sizes and enhances interpretability.

Findings

CF-T2NN outperforms state-of-the-art methods on 10 real-world datasets.

It provides more reliable uncertainty estimates in graph classification.

The method improves interpretability of GNN predictions.

Abstract

Graph Neural Networks (GNNs) have become the de facto standard for analyzing graph-structured data, leveraging message-passing techniques to capture both structural and node feature information. However, recent studies have raised concerns about the statistical reliability of uncertainty estimates produced by GNNs. This paper addresses this crucial challenge by introducing a novel technique for quantifying uncertainty in non-exchangeable graph-structured data, while simultaneously reducing the size of label prediction sets in graph classification tasks. We propose Conformalized Tensor-based Topological Neural Networks (CF-T2NN), a new approach for rigorous prediction inference over graphs. CF-T2NN employs tensor decomposition and topological knowledge learning to navigate and interpret the inherent uncertainty in decision-making processes. This method enables a more nuanced understanding and handling of prediction uncertainties, enhancing the reliability and interpretability of neural network outcomes. Our empirical validation, conducted across 10 real-world datasets, demonstrates the superiority of CF-T2NN over a wide array of state-of-the-art methods on various graph benchmarks. This work not only enhances the GNN framework with robust uncertainty quantification capabilities but also sets a new standard for reliability and precision in graph-structured data analysis.