Title: Regularized subspace clustering for functional data
Authors: Yoshikazu Terada - Osaka University; RIKEN (Japan) [presenting]
Michio Yamamoto - Osaka University / RIKEN AIP (Japan)
Abstract: The intrinsic high dimensional nature of functional data often makes possible the very good performance in supervised classification for functional data. In the supervised classification problems, it is known that, using the projection into the finite-dimensional subspace, we can extract the intrinsic high dimensional nature from functional data. In the context of unsupervised classification, there are several clustering methods based on the projection into the subspace. However, the projected data do not necessarily reflect the hidden true cluster structure in these existing methods. We propose a new regularized subspace clustering algorithm for functional data based on a cluster-separation criterion in the finite-dimensional subspace. The proposed algorithm monotonically decreases the objective function value. Moreover, we study asymptotic properties of our clustering method. The proposed method works well not only for the simulated data, but also for the real datasets which are difficult to obtain a good classification performance.