Statistical Clustering of Shape Data
Author(s):
Abstract:
Recently، employing multivariate statistical techniques for data، that are geometrically random، made more attention by the researchers from applied disciplines. Shape statistics، as a new branch of stochastic geometry، constitute batch of such data. However، due to non-Euclidean feature of such data، adopting usual tools from the multivariate statistics to proper statistical analysis of them is not somewhat clear. How to cluster the shape data is studied in this paper and then its performance is compared with the traditional view of multivariate statistics to this subject via applying these methods to analysis the distal femur.
Keywords:
Language:
Persian
Published:
Journal of Statistical Sciences, Volume:8 Issue: 2, 2014
Pages:
223 to 243
https://www.magiran.com/p1331164