Title: Spatially constrained functional clustering using nearest neighbors
Authors: Yu-Ting Chen - National Chengchi University (Taiwan) [presenting]
John Aston - University of Cambridge (United Kingdom)
Jeng-Min Chiou - Academia Sinica (Taiwan)
Abstract: Brain architecture is well known to not be simply located randomly and spatially from one brain to another, and spatial information should not be ignored in the analysis of data. Thus, simply determining clusters from their functional responses is insufficient; the spatial location from which they originate should also be considered. We proposed spatially constrained functional data clustering, motivated by the demand for it in electroencephalography (EEG) analysis based on multilevel functional principal component analysis. The objective was to cluster the EEG channels that had similar response patterns in the recorded event-related potentials.