Graph Convolutional Neural Networks Sensitivity under Probabilistic Error Model

arXiv:2203.07831v4 Announce Type: replace-cross
Abstract: Graph Neural Networks (GNNs), particularly Graph Convolutional Neural Networks (GCNNs), have emerged as pivotal instruments in machine learning and signal processing for processing graph-structured data. This paper proposes an analysis framework to investigate the sensitivity of GCNNs to probabilistic graph perturbations, directly impacting the graph shift operator (GSO). Our study establishes tight expected GSO error bounds, which are explicitly linked to the error model parameters, and reveals a linear relationship between GSO perturbations and the …

abstract analysis arxiv convolutional convolutional neural networks cs.lg data error framework gnns graph graph neural networks machine machine learning networks neural networks paper pivotal processing sensitivity shift signal stat.ml structured data the graph type