A Commentary on Noyce, Küssner and Sollich: Quantifying Shapes

Abstract

This review focuses on the statistical aspects of "Quantifying Shapes". A number of decisions made by the authors are discussed regarding the choice of methods and how exactly they are applied and their results interpreted. Most space is devoted to issues of cluster analysis and low dimensional embedding, with further remarks concerning GP regression and classification.

Keywords

• Cluster analysis
• regression
• spectral clustering
• principal components