A collection of graph classiifcation methods.
All methods are more or less modified to accept graph data from https://ls11-www.cs.tu-dortmund.de/staff/morris/graphkerneldatasets Codes are adapted from https://github.com/fanyun-sun/graph-classification and https://github.com/horacepan/MLGkernel
Download data to ./data
and refer to go.sh
under every directory for example usage.
To compile successfully, it's better to do a full compiling by doing both "make clean" and "make all"
Methods
- DGK: Deep Graph Kernels [source]
- MLGkernel: Multiscale Laplacian Graph Kernel [source]
- graph2vec_tf: graph2vec: Learning distributed representations of graphs [source]
- diffpool: Hierarchical Graph Representation Learning with Differentiable Pooling [source]
- sub2vec: Sub2Vec: Feature Learning for Subgraphs [source]
- kcnn: Kernel Graph Convolutional Neural Networks [source]
- kernel_methods: Various graph kernels implementation using GraKel