Abstract:We propose interpretable graph neural networks for sampling and recovery of graph signals, respectively. To take informative measurements, we propose a new graph neural sampling module, which aims to select those vertices that maximally express their corresponding neighborhoods. Such expressiveness can be quantified by the mutual information between vertices' features and neighborhoods' features, which are estimated via a graph neural network. To reconstruct an original graph signal from the sampled measurements, we propose a graph neural recovery module based on the algorithm-unrolling technique. Compared to previous analytical sampling and recovery, the proposed methods are able to flexibly learn a variety of graph signal models from data by leveraging the learning ability of neural networks; compared to previous neural-network-based sampling and recovery, the proposed methods are designed through exploiting specific graph properties and provide interpretability. We further design a new multiscale graph neural network, which is a trainable multiscale graph filter bank and can handle various graph-related learning tasks. The multiscale network leverages the proposed graph neural sampling and recovery modules to achieve multiscale representations of a graph. In the experiments, we illustrate the effects of the proposed graph neural sampling and recovery modules and find that the modules can flexibly adapt to various graph structures and graph signals. In the task of active-sampling-based semi-supervised learning, the graph neural sampling module improves the classification accuracy over 10% in Cora dataset. We further validate the proposed multiscale graph neural network on several standard datasets for both vertex and graph classification. The results show that our method consistently improves the classification accuracies.

Clustering-Aided Graph Signal Sampling and Reconstruction for Large-Scale Sensor Networks

Hierarchical Spatial Clustering in Multi-Hop Wireless Sensor Networks

Cluster Localization Scheme for High-Density Wireless Sensor Networks

Distributed Spatial Correlation-Based Clustering for Approximate Data Collection in Wsns

Data Aggregation Scheme Based on Multi-Resolution and Compressive Sensing in Wireless Sensor Network

A Graph Embedding Method for Wireless Sensor Networks Localization

Graph-Based Signal Sampling with Adaptive Subspace Reconstruction for Spatially-Irregular Sensor Data

Data Collection Algorithm for Clustered Large-Scale Wireless Sensor Networks Using Virtual Nodes

Reconstruction of Bandlimited Graph Signals from Random Local Sampling

Generalized Graph Signal Sampling And Reconstruction

Cluster-preserving Sampling Algorithm for Large-Scale Graphs.

A Distributed Tracking Algorithm for Reconstruction of Graph Signals

Compressive Sensing Based Sampling and Reconstruction for Wireless Sensor Array Network

Sampling and Recovery of Graph Signals based on Graph Neural Networks

Clustering Affiliation Inference from Graph Samples

Recovery of Missing Sensor Data by Reconstructing Time-varying Graph Signals

Data Reconstruction Coverage Based on Graph Signal Processing for Wireless Sensor Networks

Sensor Network Localization Using Kernel Spectral Regression

Reconstructing Graph Signals from Noisy Dynamical Samples

Optimized Random Deployment of Large-Scale Energy-Harvesting Sensors for Field Reconstruction

Graph-based Reconstruction of Time-Varying Spatial Signals