The library provides functions that allow:
- Build the affinity matrix.
- Perform the embedding of the patterns in the space spanned by the eigenvectors of the matrices derived from the affinity matrix.
- Obtain an approximation of the eigenvectors in order to reduce the computational complexity.
- Exploiting information from multiple views. Corresponding nodes in each graph should have the same cluster membership.
- Clusterize the eigenvector space.
- Graph construction
- Approximate embedding
- Multiple views
- Kernel Addition
- Kernel Product
- Feature Concatenation (in the examples section)
- Co-regularized Multi-view Spectral Clustering
The documentation and the library is still a work in progress.