Fast Neighborhood Subgraph Pairwise Distance Kernel

We introduce a novel graph kernel called the Neighborhood Subgraph Pairwise Distance Kernel. The kernel decomposes a graph into all pairs of neighborhood subgraphs of small radius at increasing distances. We show that using a fast graph invariant we obtain significant speed-ups in the Gram matrix computation. Finally, we test the novel kernel on a wide range of chemoinformatics tasks, from antiviral to anticarcinogenic to toxicological activity prediction, and observe competitive performance when compared against several recent graph kernel methods.

• Type: Conference
• Year: 2010
• Where: ICML
• Authors: Fabrizio Costa, Kurt De Grave
• Keywords: Graph Kernel, ICML 2010, Machine Learning, Neighborhood Subgraphs, Pairwise Distance Kernel