B0858
Title: Extending barycentric subspace analysis to a set of graphs
Authors: Elodie Maignant - INRIA (France)
Xavier Pennec - Université Côte d'Azur and INRIA (France)
Anna Calissano - Imperial College London (Italy) [presenting]
Abstract: Barycentric subspace analysis (BSA) is introduced for a set of graphs. Identifying each graph by its eigenvalues set, the graph spectrum space is built, a quotient manifold of isospectral graphs. In such a manifold, the notion of BSA is extended. It showcases how BSA can be used as a powerful dimensionality reduction technique for complex data. BSA searches for a subspace of a lower dimension, minimizing the projection of data points on such subspace. As the subspace is identified by a set of reference points (which are data points if constrained to the data), the interpretation is easier than with other dimensionality techniques. BSA is performed and is compared with clustering and PCA on a simulated dataset and a real-world dataset of airline company networks.
