Spatial Graph Convolution Neural Networks for Water Distribution Systems
This work addresses a critical infrastructure problem for water distribution systems, offering incremental improvements in graph neural network methods for this domain.
The paper tackles missing value estimation in water distribution system graphs by proposing a spatial graph convolution neural network architecture that addresses long-term dependencies, achieving excellent results on benchmark tasks and introducing a multi-hop variation that reduces resource requirements for large graphs.
We investigate the task of missing value estimation in graphs as given by water distribution systems (WDS) based on sparse signals as a representative machine learning challenge in the domain of critical infrastructure. The underlying graphs have a comparably low node degree and high diameter, while information in the graph is globally relevant, hence graph neural networks face the challenge of long-term dependencies. We propose a specific architecture based on message passing which displays excellent results for a number of benchmark tasks in the WDS domain. Further, we investigate a multi-hop variation, which requires considerably less resources and opens an avenue towards big WDS graphs.