Graph signal recovery using restricted Boltzmann machines

作者：

Highlights：

• We propose a model-agnostic pipeline to recover graph signals using RBMs and neural networks.

• Denoising representations learned by our model is more effective than denoising the data itself.

• We establish a relationship between the clustering of the representations and prediction accuracy.

摘要

•We propose a model-agnostic pipeline to recover graph signals using RBMs and neural networks.•Denoising representations learned by our model is more effective than denoising the data itself.•We establish a relationship between the clustering of the representations and prediction accuracy.

论文关键词：Graph signal recovery,Noisy data problem,Restricted Boltzmann machines,Deep neural networks,Graph neural networks,Social network analysis,Denoising models

论文评审过程：Received 20 November 2020, Revised 13 April 2021, Accepted 18 July 2021, Available online 30 July 2021, Version of Record 4 August 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.115635