Hi! Does it make sense to use the Multi-Response Permutation Procedure (MRPP) for an evaluation of k-means clustering runs, e.g. as a substitute for the Bayesian information criterion?
Thank you very much in advance, Carl . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
