Graph Signal Recovery Using Restricted Boltzmann Machines

TL;DR

This paper introduces a novel graph signal recovery method leveraging restricted Boltzmann machines and neural networks, demonstrating improved denoising effectiveness especially for graph-structured data.

Contribution

The paper presents a model-agnostic pipeline combining RBMs and neural networks for effective graph signal recovery from noisy data.

Findings

Denoising learned representations outperforms data denoising.

Pipeline is particularly effective for graph-structured datasets.

Model-agnostic approach adaptable to various data types.

Abstract

We propose a model-agnostic pipeline to recover graph signals from an expert system by exploiting the content addressable memory property of restricted Boltzmann machine and the representational ability of a neural network. The proposed pipeline requires the deep neural network that is trained on a downward machine learning task with clean data, data which is free from any form of corruption or incompletion. We show that denoising the representations learned by the deep neural networks is usually more effective than denoising the data itself. Although this pipeline can deal with noise in any dataset, it is particularly effective for graph-structured datasets.