Graph Diffusion & PCA Framework for Semi-supervised Learning - Archive ouverte HAL Access content directly
Conference Papers Year : 2021

Graph Diffusion & PCA Framework for Semi-supervised Learning

(1) , (2) , (1, 2)
1
2

Abstract

A novel framework called Graph Diffusion & PCA (GDPCA) is proposed in the context of semi-supervised learning on graph structured data. It combines a modified Principal Component Analysis with the classical supervised loss and Laplacian regularization, thus handling the case where the adjacency matrix is sparse and avoiding the curse of dimensionality. Our framework can be applied to non-graph datasets as well, such as images by constructing similarity graph. GDPCA improves node classification by enriching the local graph structure by node covariance. We demonstrate the performance of GDPCA in experiments on citation networks and images, and we show that GDPCA compares favourably with the best state-of-the-art algorithms and has significantly lower computational complexity.
Fichier principal
Vignette du fichier
LION2021_author_version.pdf (1.03 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03477308 , version 1 (13-12-2021)

Identifiers

Cite

Konstantin Avrachenkov, Aurélie Boisbunon, Mikhail Kamalov. Graph Diffusion & PCA Framework for Semi-supervised Learning. LION 2021 - 15th Learning and Intelligent Optimization Conference, Jun 2021, Athens, Greece. pp.25-39, ⟨10.1007/978-3-030-92121-7_3⟩. ⟨hal-03477308⟩
41 View
105 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More