Title: A novel approach to joint sparse functional clustering and alignment
Authors: Valeria Vitelli - University of Oslo (Norway) [presenting]
Abstract: When performing functional clustering, the problem of selecting the portions of the domain which are most relevant to the classification purposes has already been considered. When misalignment is also present, the only possible approach is to first align the curves, and then use a sparse functional clustering method to estimate the groups and select the domain. However, it has been already proved that aligning and clustering the curves jointly is beneficial for the analysis. We thus propose a novel algorithm which jointly performs all these tasks: clustering, alignment, and domain selection. We prove the well-posedness of the problem, and test the method on simulated data. We also perform the analysis of the Berkeley Growth Study data, as a benchmark for functional data, and propose the use of the method for genomic data.