Simplifying GNN Performance with Low Rank Kernel Models

Luciano Vinas, Arash A. Amini

Primary Area: learning on graphs and other geometries & topologies

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Keywords: GNN, Spectral filtering, Semi-supervised node classification, Kernel methods, Low rank

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2024/AuthorGuide.

TL;DR: Many of the current GNN architectures may be over-engineered for small graph semi-supervised node classification and may instead be replaced by simpler, traditional methods from nonparametric estimation, applied in the graph spectral domain.

Abstract:

We revisit recent spectral GNN approaches to
semi-supervised node classification (SSNC).
We posit that many of the current GNN architectures may be over-engineered. Instead, simpler, traditional methods from nonparametric estimation, applied in the spectral domain, could replace many deep-learning inspired GNN designs. These conventional techniques appear to be well suited for a variety of graph types reaching state-of-the-art performance on many the common SSNC benchmarks. Additionally, we show that recent performance improvements in GNN approaches may be partialy attributed to shifts in evaluation conventions. Lastly, an ablative study is conducted on the various hyperparameters associated with GNN spectral filtering techniques.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors' identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 8222