Outlier identification with the Harmonic Topographic Mapping

We review two versions of a topology preserving algorithm one of which we had previously [1] found to be more successful in defining smooth manifolds and tight clusters. In the context of outlier detection, however, the other is shown to be more successful. Nevertheless, we show that, by using local kernels for calculation of responsibilities, the first one can also be used in this manner.

• Type: Conference
• Year: 2006
• Where: ESANN
• Authors: Marian Pena, Colin Fyfe
• Keywords: ESANN 2006, ESANN 2007, Outlier Detection, Smooth Manifolds, Topology Preserving Algorithm