Assignment for the course of Artificial Intelligence: Knowledge Representation and Planning, taught by Professor Andrea Torsello of the Ca' Foscari University of Venice.
Check the report for a complete analysis of the task and much more information on the actual implementation and the results.
Read this article presenting a way to improve the disciminative power of graph kernels.
Choose one graph kernel among
- Shortest-path Kernel
- Graphlet Kernel
- Random Walk Kernel
- Weisfeiler-Lehman Kernel
Choose one manifold learning technique among
- Isomap
- Diffusion Maps
- Laplacian Eigenmaps
- Local Linear Embedding
Compare the performance of an SVM trained on the given kernel, with or without the manifold learning step, on the following datasets:
Note: the datasets are contained in Matlab files. The variable G contains a vector of cells, one per graph. The entry am of each cell is the adjacency matrix of the graph. The variable labels, contains the class-labels of each graph.