Clustering for high-dimension, low-sample size data using distance vectors