Generalization in Graph Neural Networks: Improved PAC-Bayesian Bounds on Graph Diffusion

• URL: http://arxiv.org/abs/2302.04451v3
• Date: Mon, 23 Oct 2023 21:37:55 GMT
• Title: Generalization in Graph Neural Networks: Improved PAC-Bayesian Bounds on Graph Diffusion
• Authors: Haotian Ju, Dongyue Li, Aneesh Sharma, and Hongyang R. Zhang
• Abstract summary: We present generalization bounds that scale with the largest singular value of the graph neural network's feature diffusion matrix. These bounds are numerically much smaller than prior bounds for real-world graphs. We measure the stability of graph neural networks against noise perturbations using Hessians.
• Score: 17.70434437597516
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Graph neural networks are widely used tools for graph prediction tasks. Motivated by their empirical performance, prior works have developed generalization bounds for graph neural networks, which scale with graph structures in terms of the maximum degree. In this paper, we present generalization bounds that instead scale with the largest singular value of the graph neural network's feature diffusion matrix. These bounds are numerically much smaller than prior bounds for real-world graphs. We also construct a lower bound of the generalization gap that matches our upper bound asymptotically. To achieve these results, we analyze a unified model that includes prior works' settings (i.e., convolutional and message-passing networks) and new settings (i.e., graph isomorphism networks). Our key idea is to measure the stability of graph neural networks against noise perturbations using Hessians. Empirically, we find that Hessian-based measurements correlate with the observed generalization gaps of graph neural networks accurately. Optimizing noise stability properties for fine-tuning pretrained graph neural networks also improves test performance on several graph-level classification tasks.

Related papers

• Generalization Error of Graph Neural Networks in the Mean-field Regime [10.35214360391282]
  We explore two widely utilized types of graph neural networks: graph convolutional neural networks and message passing graph neural networks. Our novel approach involves deriving upper bounds within the mean-field regime for evaluating the generalization error of these graph neural networks.
  arXiv  Detail & Related papers  (2024-02-10T19:12:31Z)
• Learning Graph Structure from Convolutional Mixtures [119.45320143101381]
  We propose a graph convolutional relationship between the observed and latent graphs, and formulate the graph learning task as a network inverse (deconvolution) problem. In lieu of eigendecomposition-based spectral methods, we unroll and truncate proximal gradient iterations to arrive at a parameterized neural network architecture that we call a Graph Deconvolution Network (GDN) GDNs can learn a distribution of graphs in a supervised fashion, perform link prediction or edge-weight regression tasks by adapting the loss function, and they are inherently inductive.
  arXiv  Detail & Related papers  (2022-05-19T14:08:15Z)
• Optimal Propagation for Graph Neural Networks [51.08426265813481]
  We propose a bi-level optimization approach for learning the optimal graph structure. We also explore a low-rank approximation model for further reducing the time complexity.
  arXiv  Detail & Related papers  (2022-05-06T03:37:00Z)
• SpeqNets: Sparsity-aware Permutation-equivariant Graph Networks [16.14454388348814]
  We present a class of universal, permutation-equivariant graph networks. They offer a fine-grained control between expressivity and scalability and adapt to the sparsity of the graph. These architectures lead to vastly reduced computation times compared to standard higher-order graph networks.
  arXiv  Detail & Related papers  (2022-03-25T21:17:09Z)
• Hyperbolic Graph Neural Networks: A Review of Methods and Applications [55.5502008501764]
  Graph neural networks generalize conventional neural networks to graph-structured data. The performance of Euclidean models in graph-related learning is still bounded and limited by the representation ability of Euclidean geometry. Recently, hyperbolic space has gained increasing popularity in processing graph data with tree-like structure and power-law distribution.
  arXiv  Detail & Related papers  (2022-02-28T15:08:48Z)
• How Neural Processes Improve Graph Link Prediction [35.652234989200956]
  We propose a meta-learning approach with graph neural networks for link prediction: Neural Processes for Graph Neural Networks (NPGNN) NPGNN can perform both transductive and inductive learning tasks and adapt to patterns in a large new graph after training with a small subgraph.
  arXiv  Detail & Related papers  (2021-09-30T07:35:13Z)
• Increase and Conquer: Training Graph Neural Networks on Growing Graphs [116.03137405192356]
  We consider the problem of learning a graphon neural network (WNN) by training GNNs on graphs sampled Bernoulli from the graphon. Inspired by these results, we propose an algorithm to learn GNNs on large-scale graphs that, starting from a moderate number of nodes, successively increases the size of the graph during training.
  arXiv  Detail & Related papers  (2021-06-07T15:05:59Z)
• Towards Deeper Graph Neural Networks [63.46470695525957]
  Graph convolutions perform neighborhood aggregation and represent one of the most important graph operations. Several recent studies attribute this performance deterioration to the over-smoothing issue. We propose Deep Adaptive Graph Neural Network (DAGNN) to adaptively incorporate information from large receptive fields.
  arXiv  Detail & Related papers  (2020-07-18T01:11:14Z)
• Graph Structure of Neural Networks [104.33754950606298]
  We show how the graph structure of neural networks affect their predictive performance. A "sweet spot" of relational graphs leads to neural networks with significantly improved predictive performance. Top-performing neural networks have graph structure surprisingly similar to those of real biological neural networks.
  arXiv  Detail & Related papers  (2020-07-13T17:59:31Z)
• A Note on Over-Smoothing for Graph Neural Networks [13.008323851750442]
  We build upon previous results citeoono 2019graph to analyze the over-smoothing effect in the general graph neural network architecture. We show when the weight matrix satisfies the conditions determined by the spectrum of augmented normalized Laplacian, the Dirichlet energy of embeddings will converge to zero, resulting in the loss of discriminative power.
  arXiv  Detail & Related papers  (2020-06-23T20:36:56Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.