Graph Classification Gaussian Processes via Hodgelet Spectral Features
This work addresses graph classification for machine learning practitioners, offering a method to handle edge features, which is incremental over existing vertex-only approaches.
The authors tackled the problem of graph classification by developing a Gaussian process-based algorithm that leverages both vertex and edge features, using Hodge decomposition to capture their richness, resulting in improved performance on diverse tasks.
The problem of classifying graphs is ubiquitous in machine learning. While it is standard to apply graph neural networks or graph kernel methods, Gaussian processes can be employed by transforming spatial features from the graph domain into spectral features in the Euclidean domain, and using them as the input points of classical kernels. However, this approach currently only takes into account features on vertices, whereas some graph datasets also support features on edges. In this work, we present a Gaussian process-based classification algorithm that can leverage one or both vertex and edges features. Furthermore, we take advantage of the Hodge decomposition to better capture the intricate richness of vertex and edge features, which can be beneficial on diverse tasks.