Abdullah Shaik

Graph Classification via Network Usable Information: From Representation Evaluation to Structure Selection

Abdullah Shaik, Anwar Said

We propose NetinfoGC, a framework for graph classification that extends the Network Usable Information (NUI) paradigm to graph-level learning. Unlike conventional graph neural network approaches that rely on end-to-end training of black-box embeddings, NetinfoGC constructs a fami…Read more