A soft k-segments algorithm for principal curves
Abstract
We propose a new method to find principal curves for data sets. The method repeats three steps until a stopping criterion is met. In the first step, k (unconnected) line segments are fitted on the data. The second step connects the segments to form a polygonal line, and evaluates the quality of the resulting polygonal line. The third step inserts a new line segment. We compare the performance of our new method with other existing methods to find principal curves.
Domains
Machine Learning [cs.LG]
Fichier principal
verbeek01icann2.pdf (180.2 Ko)
Télécharger le fichier
VVK01.png (13.13 Ko)
Télécharger le fichier
Origin : Files produced by the author(s)
Format : Figure, Image