
A Hierarchical TransitiveAligned Graph Kernel for Unattributed Graphs
In this paper, we develop a new graph kernel, namely the Hierarchical Tr...
read it

Graph Kernels via Functional Embedding
We propose a representation of graph as a functional object derived from...
read it

The Multiscale Laplacian Graph Kernel
Many real world graphs, such as the graphs of molecules, exhibit structu...
read it

Density of States Graph Kernels
An important problem on graphstructured data is that of quantifying sim...
read it

Towards a practical kdimensional WeisfeilerLeman algorithm
The kdimensional WeisfeilerLeman algorithm is a wellknown heuristic f...
read it

Neighborhood Preserving Kernels for Attributed Graphs
We describe the design of a reproducing kernel suitable for attributed g...
read it

A Generalized WeisfeilerLehman Graph Kernel
The WeisfeilerLehman graph kernels are among the most prevalent graph k...
read it
Global WeisfeilerLehman Graph Kernels
Most stateoftheart graph kernels only take local graph properties into account, i.e., the kernel is computed with regard to properties of the neighborhood of vertices or other small substructures. On the other hand, kernels that do take global graph propertiesinto account may not scale well to large graph databases. Here we propose to start exploring the space between local and global graph kernels, striking the balance between both worlds. Specifically, we introduce a novel graph kernel based on the kdimensional WeisfeilerLehman algorithm. Unfortunately, the kdimensional WeisfeilerLehman algorithm scales exponentially in k. Consequently, we devise a stochastic version of the kernel with provable approximation guarantees using conditional Rademacher averages. On boundeddegree graphs, it can even be computed in constant time. We support our theoretical results with experiments on several graph classification benchmarks, showing that our kernels often outperform the stateoftheart in terms of classification accuracies.
READ FULL TEXT
Comments
There are no comments yet.