On the Expressive Power of Sparse Geometric MPNNs

• URL: http://arxiv.org/abs/2407.02025v2
• Date: Wed, 09 Oct 2024 06:47:38 GMT
• Title: On the Expressive Power of Sparse Geometric MPNNs
• Authors: Yonatan Sverdlov, Nadav Dym,
• Abstract summary: We study the expressive power of message-passing neural networks for geometric graphs. We show that generic pairs of non-isomorphic geometric graphs can be separated by message-passing networks.
• Score: 3.396731589928944
• License:
• Abstract: Motivated by applications in chemistry and other sciences, we study the expressive power of message-passing neural networks for geometric graphs, whose node features correspond to 3-dimensional positions. Recent work has shown that such models can separate \emph{generic} pairs of non-isomorphic geometric graphs, though they may fail to separate some rare and complicated instances. However, these results assume a fully connected graph, where each node possesses complete knowledge of all other nodes. In contrast, often, in application, every node only possesses knowledge of a small number of nearest neighbors. This paper shows that generic pairs of non-isomorphic geometric graphs can be separated by message-passing networks with rotation equivariant features as long as the underlying graph is connected. When only invariant intermediate features are allowed, generic separation is guaranteed for generically globally rigid graphs. We introduce a simple architecture, $\us$, which achieves our theoretical guarantees and compares favorably with alternative architecture on synthetic and chemical benchmarks. Our code is available at \url{https://github.com/yonatansverdlov/E-GenNet}.

Related papers

• A Hitchhiker's Guide to Geometric GNNs for 3D Atomic Systems [87.30652640973317]
  Recent advances in computational modelling of atomic systems represent them as geometric graphs with atoms embedded as nodes in 3D Euclidean space. Geometric Graph Neural Networks have emerged as the preferred machine learning architecture powering applications ranging from protein structure prediction to molecular simulations and material generation. This paper provides a comprehensive and self-contained overview of the field of Geometric GNNs for 3D atomic systems.
  arXiv  Detail & Related papers  (2023-12-12T18:44:19Z)
• NodeFormer: A Scalable Graph Structure Learning Transformer for Node Classification [70.51126383984555]
  We introduce a novel all-pair message passing scheme for efficiently propagating node signals between arbitrary nodes. The efficient computation is enabled by a kernerlized Gumbel-Softmax operator. Experiments demonstrate the promising efficacy of the method in various tasks including node classification on graphs.
  arXiv  Detail & Related papers  (2023-06-14T09:21:15Z)
• Seq-HGNN: Learning Sequential Node Representation on Heterogeneous Graph [57.2953563124339]
  We propose a novel heterogeneous graph neural network with sequential node representation, namely Seq-HGNN. We conduct extensive experiments on four widely used datasets from Heterogeneous Graph Benchmark (HGB) and Open Graph Benchmark (OGB)
  arXiv  Detail & Related papers  (2023-05-18T07:27:18Z)
• GrannGAN: Graph annotation generative adversarial networks [72.66289932625742]
  We consider the problem of modelling high-dimensional distributions and generating new examples of data with complex relational feature structure coherent with a graph skeleton. The model we propose tackles the problem of generating the data features constrained by the specific graph structure of each data point by splitting the task into two phases. In the first it models the distribution of features associated with the nodes of the given graph, in the second it complements the edge features conditionally on the node features.
  arXiv  Detail & Related papers  (2022-12-01T11:49:07Z)
• Break the Wall Between Homophily and Heterophily for Graph Representation Learning [25.445073413243925]
  Homophily and heterophily are intrinsic properties of graphs that describe whether two linked nodes share similar properties. This work identifies three graph features, including the ego node feature, the aggregated node feature, and the graph structure feature, that are essential for graph representation learning. It proposes a new GNN model called OGNN that extracts all three graph features and adaptively fuses them to achieve generalizability across the whole spectrum of homophily.
  arXiv  Detail & Related papers  (2022-10-08T19:37:03Z)
• SGAT: Simplicial Graph Attention Network [38.7842803074593]
  Heterogeneous graphs have multiple node and edge types and are semantically richer than homogeneous graphs. Many graph neural network approaches for heterogeneous graphs use metapaths to capture multi-hop interactions between nodes. We present Simplicial Graph Attention Network (SGAT), a simplicial complex approach to represent such high-order interactions.
  arXiv  Detail & Related papers  (2022-07-24T15:20:41Z)
• Graph Neural Networks with Learnable Structural and Positional Representations [83.24058411666483]
  A major issue with arbitrary graphs is the absence of canonical positional information of nodes. We introduce Positional nodes (PE) of nodes, and inject it into the input layer, like in Transformers. We observe a performance increase for molecular datasets, from 2.87% up to 64.14% when considering learnable PE for both GNN classes.
  arXiv  Detail & Related papers  (2021-10-15T05:59:15Z)
• Factorizable Graph Convolutional Networks [90.59836684458905]
  We introduce a novel graph convolutional network (GCN) that explicitly disentangles intertwined relations encoded in a graph. FactorGCN takes a simple graph as input, and disentangles it into several factorized graphs. We evaluate the proposed FactorGCN both qualitatively and quantitatively on the synthetic and real-world datasets.
  arXiv  Detail & Related papers  (2020-10-12T03:01:40Z)
• Geom-GCN: Geometric Graph Convolutional Networks [15.783571061254847]
  We propose a novel geometric aggregation scheme for graph neural networks to overcome the two weaknesses. The proposed aggregation scheme is permutation-invariant and consists of three modules, node embedding, structural neighborhood, and bi-level aggregation. We also present an implementation of the scheme in graph convolutional networks, termed Geom-GCN, to perform transductive learning on graphs.
  arXiv  Detail & Related papers  (2020-02-13T00:03:09Z)

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.