Geosphere; February 2008; v. 4; no. 1;
p. 247-259; DOI: 10.1130/GES00139.1
© Geological Society of America
Click on image to view larger version.
Figure 1. Surface classification algorithm applied to synthetic data. (A) Pass 1, the multileader cluster algorithm produces 81 clusters with an average (Navg) of 15 points per cluster (where Nmin is the minimum number of points allowed in each cluster and Nmax is the maximum number of points allowed per cluster). (B) Pass 2, the final partition produced by the k-means cluster algorithm identified three clusters, corresponding to the three orientations of the six faces of the test shape. Parallel faces were assigned to the same cluster since they have the same orientation. (C) The optimum number of three clusters in the final partition is determined by the L-shaped break value of the minimum description length (MDL) criterion. (D) Equal-area stereographic projection of the orientation data. The distribution of poles to the planes defined in the pass 1 partition indicates that there is a strong clustering about the mean for each orientation. Mean vectors for each cluster are plotted as circles.
Copyright © 2008 by Geological Society of America
