Abstract: Graphs are effective tools for modeling complex data. Setting out from two basic substructures, random walks and trees, we propose a new family of context-dependent random walk graph kernels ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results