109
click to vote
IDEAL
2000
Springer
15 years 6 months ago
2000
Springer
Abstract. We present a clustering method for continuous data. It defines local clusters into the (primary) data space but derives its similarity measure from the posterior distribu...