Flexible and Robust Bayesian Classification by Finite Mixture Models
(2004) ESANN 2004, European Symposium on Artificial Neural Networks — Location: Bruges (Belgium) (28.April.2004)
Files
94-FlexibleandRobustBayesianClassificationbyFiniteMixtureModels.pdf
Restricted Access - Adobe PDF
- 146.44 KB
Details
- Authors
- Abstract
- The regularized Mahalanobis distance is proposed in the framework of 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 classication is near optimal on Ripley's synthetic data set.
- Affiliations
Citations
Archambeau, C., Vrins, F., & Verleysen, M. (2004). Flexible and Robust Bayesian Classification by Finite Mixture Models. Proceedings of ESANN 2004, European Symposium on Artificial Neural Networks, p. 75-80. https://hdl.handle.net/2078.5/223048