An EM Stopping Rule for Avoiding Degeneracy in Gaussian-Based Clustering withMissing Data.

Missing data frequency increases with the growing size of multivariate moderndatasets. In Gaussian model-based clustering, the EM algorithm easily takes into account suchdata but the degeneracy problem is dramatically aggravated during the EM runs: parameter de-generacy is quite slow and also more frequent than with complete data. Consequently, param-eter degenerated solutions may be confused with valuable parameter solutions and, in addition,computing time may be wasted through wrong runs. In this work, a simple and low informational condition on the latent partition allows to propose a very simple partition-based stoppingrule of EM which shows good behavior on nu- merical experiments.