A1399
Title: Changepoint detection in tensor data
Authors: Barbora Pestova - Charles University (Czech Republic) [presenting]
Michal Pesta - Charles University (Czech Republic)
Martin Romanak - Charles University (Czech Republic)
Abstract: Tensor data consisting of multivariate outcomes over the items and across the subjects with longitudinal and cross-sectional dependence are considered. Distributional-free detection procedures for changepoints at different locations are proposed, which are in an unsupervised learning manner. The bootstrap superstructure is developed to overcome computational issues in such a universal setup. The completely data-driven test is presented using real data examples.
