Principal curve classifier-a nonlinear approach to pattern classification

Presents a general nonlinear approach to pattern classification using principal curves. Principal curves are nonparametric, nonlinear generalizations of the first principal component, and may also be regarded as continuous versions of 1-D self-organizing maps. The new classification technique, principal curve classifier (PCC), involves a novel way of computing a principal curve for each class using the class-labeled training data. An unlabeled test point is given the class-label of the principal curve that is closest to it in Euclidean distance. Preliminary experiments comparing the PCC with established classification methods, using selected datasets from the Elena and Proben1 benchmarks, highlight the merits and limitations of this algorithm