Flexible and Robust Bayesian Classification by Finite Mixture Models.
Presented at: UNSPECIFIED.
The regularized Mahalanobis distance is proposed in the framework of fi nite mixture models to avoid commonly faced numerical difficulties encountered with EM. Its principle is applied to Gaussian and Student-t mixtures, resulting in reliable density estimates, the model complexity being kept low. Besides, the regularized models are robust to various noise types. Finally, it is shown that the quality of the associated Bayesian classifi cation is near optimal on Ripley's synthetic data set.
|Type:||Conference item (UNSPECIFIED)|
|Title:||Flexible and Robust Bayesian Classification by Finite Mixture Models|
|UCL classification:||UCL > School of BEAMS > Faculty of Maths and Physical Sciences
UCL > School of BEAMS > Faculty of Maths and Physical Sciences > Statistical Science
Archive Staff Only