An optimization-based approach to node role discovery in networks: approximating equitable partitions

M Scholkemper, MT Schaub - Advances in Neural …, 2024 - proceedings.neurips.cc
Similar to community detection, partitioning the nodes of a complex network according to
their structural roles aims to identify fundamental building blocks of a network, which can be …

Multi-scale Wasserstein Shortest-path Graph Kernels for Graph Classification

W Ye, H Tian, Q Chen - IEEE Transactions on Artificial …, 2023 - ieeexplore.ieee.org
Graph kernels are conventional methods for computing graph similarities. However, the
existing R-convolution graph kernels cannot resolve both of the two challenges: 1) …

Exploring Consistency in Graph Representations: from Graph Kernels to Graph Neural Networks

X Liu, Y Cai, Q Yang, Y Yan - arxiv preprint arxiv:2410.23748, 2024 - arxiv.org
Graph Neural Networks (GNNs) have emerged as a dominant approach in graph
representation learning, yet they often struggle to capture consistent similarity relationships …

Deep Hierarchical Graph Alignment Kernels

S Tang, H Tian, X Cao, W Ye - arxiv preprint arxiv:2405.05545, 2024 - arxiv.org
Typical R-convolution graph kernels invoke the kernel functions that decompose graphs into
non-isomorphic substructures and compare them. However, overlooking implicit similarities …

Towards Subgraph Isomorphism Counting with Graph Kernels

X Liu, W Wang, J Bai, Y Song - arxiv preprint arxiv:2405.07497, 2024 - arxiv.org
Subgraph isomorphism counting is known as# P-complete and requires exponential time to
find the accurate solution. Utilizing representation learning has been shown as a promising …

Enhancing Shortest-Path Graph Kernels via Graph Augmentation

W Ye, H Tian, S Tang, X Sun - Joint European Conference on Machine …, 2024 - Springer
The conventional shortest-path graph kernel (SP) decomposes graphs into shortest paths
and computes their frequencies in each graph. However, SP cannot compare graphs with …

Utilizing Constrained Homomorphisms in the Design of Efficient Graph Kernels

TH Schulz - 2024 - bonndoc.ulb.uni-bonn.de
Learning on graphs, particularly graph classification, requires rich graph representations. A
common paradigm to obtain these is by extracting sets of substructures and representing …