Currently, I am working on the application of GNNs for combinatorial optimization. In particular, how to select an optimal subgraph w.r.t. multiple graph tasks given a capacity constraint. In general, I am interested in the intersection of mathematical optimization and machine learning.