Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

NetLSD: Hearing the Shape of a Graph

About

Comparison among graphs is ubiquitous in graph analytics. However, it is a hard task in terms of the expressiveness of the employed similarity measure and the efficiency of its computation. Ideally, graph comparison should be invariant to the order of nodes and the sizes of compared graphs, adaptive to the scale of graph patterns, and scalable. Unfortunately, these properties have not been addressed together. Graph comparisons still rely on direct approaches, graph kernels, or representation-based methods, which are all inefficient and impractical for large graph collections. In this paper, we propose the Network Laplacian Spectral Descriptor (NetLSD): the first, to our knowledge, permutation- and size-invariant, scale-adaptive, and efficiently computable graph representation method that allows for straightforward comparisons of large graphs. NetLSD extracts a compact signature that inherits the formal properties of the Laplacian spectrum, specifically its heat or wave kernel; thus, it hears the shape of a graph. Our evaluation on a variety of real-world graphs demonstrates that it outperforms previous works in both expressiveness and efficiency.

Anton Tsitsulin, Davide Mottin, Panagiotis Karras, Alex Bronstein, Emmanuel M\"uller• 2018

Related benchmarks

TaskDatasetResultRank
Graph Distribution Classification and ClusteringER, RP, SBM, RG graph distributions
Accuracy90
31
Graph classification (trajectory- vs cluster-like)Single-cell graphs All Graphs (full set)
Accuracy73.1
28
Graph classification (trajectory- vs cluster-like)Single-cell graphs Gold subset
Accuracy82.3
27
Graph ExpressivityBREC
Basic Expressivity Score60
26
Graph Parameter ClassificationStochastic block model (SBM)
Accuracy83
8
Graph Distribution ClassificationRandom Graph Models ER, RP, RG, SBM
Accuracy90
8
Graph Parameter ClassificationRandom Partition (RP)
Accuracy97
8
Graph Parameter ClassificationRandom Geometric (RG) Graph
Accuracy92
8
Graph Parameter ClassificationErdős-Rényi (ER) random graph model
Accuracy97
8
Binary Graph ClassificationAll 169 Graphs (5-fold stratified CV)
Accuracy (Test)73.1
6
Showing 10 of 11 rows

Other info

Follow for update