Search Results for author:
Found 2 papers, 0 papers with code
Highly Accurate Disease Diagnosis and Highly Reproducible Biomarker Identification with PathFormer
However, we found two major limitations of existing GNNs in omics data analysis, i. e., limited-prediction (diagnosis) accuracy and limited-reproducible biomarker identification capacity across multiple datasets.
Rethinking the Power of Graph Canonization in Graph Representation Learning with Stability
We theoretically reveal the trade-off of expressivity and stability in graph-canonization-enhanced GNNs.
Cannot find the paper you are looking for? You can
Submit a new open access paper.
