Fuzzy C-means: Differences on Clustering Behavior between High Dimensional and Functional Data (Student Abstract)

Fuzzy c-means (FCM) is a generalization of the classical k-means clustering algorithm to case where an observation can belong several clusters at same time. The was previously observed have initialization problems when number desired or dimensions data are high. We tested FCM against with functional data, generated from stationary Gaussian processes, and thus in principle infinite-dimensional. that more nature, which be obtained by tuning length-scale parameter process, aforementioned do not appear. This only indicates suitable as method for but also illustrates how differs traditional multivariate data. In addition this seems suggest qualitative way measure latent dimensionality distribution itself.