ICML
2008
IEEE
14 years 12 months ago
2008
IEEE
We present a principled Bayesian framework for modeling partial memberships of data points to clusters. Unlike a standard mixture model which assumes that each data point belongs ...