Nonlinear Higher-Order Label Spreading
Tudisco, Francesco, Benson, Austin R., Prokopchik, Konstantin
Label spreading is a general technique for semi-supervised learning with point cloud or network data, which can be interpreted as a diffusion of labels on a graph. While there are many variants of label spreading, nearly all of them are linear models, where the incoming information to a node is a weighted sum of information from neighboring nodes. Here, we add nonlinearity to label spreading through nonlinear functions of higher-order structure in the graph, namely triangles in the graph. For a broad class of nonlinear functions, we prove convergence of our nonlinear higher-order label spreading algorithm to the global solution of a constrained semi-supervised loss function. We demonstrate the efficiency and efficacy of our approach on a variety of point cloud and network datasets, where the nonlinear higher-order model compares favorably to classical label spreading, as well as hypergraph models and graph neural networks.
Jun-8-2020
- Country:
- North America > United States
- New York > Tompkins County > Ithaca (0.04)
- Europe > Italy
- Abruzzo > L'Aquila Province > L'Aquila (0.04)
- North America > United States
- Genre:
- Research Report (0.82)
- Industry:
- Information Technology > Networks (0.34)
- Technology: