000082520 001__ 82520
000082520 005__ 20181203035714.0
000082520 037__ $$aCONF
000082520 245__ $$aA comparison of mixture models for density estimation
000082520 269__ $$a1999
000082520 260__ $$bLondon: IEE$$c1999
000082520 336__ $$aConference Papers
000082520 500__ $$a(IDIAP-RR 98-14)
000082520 520__ $$aGaussian mixture models (GMMs) are a popular tool for density estimation. However, these models are limited by the fact that they either impose strong constraints on the covariance matrices of the component densities or no constraints at all. This paper presents an experimental comparison of GMMs and the recently introduced mixtures of linear latent variable models. It is shown that the latter models are a more flexible alternative for GMMs and often lead to improved results.
000082520 6531_ $$alearning
000082520 6531_ $$aperry
000082520 700__ $$aMoerland, Perry
000082520 7112_ $$aProceedings of the International Conference on Artificial Neural Networks (ICANN'99)
000082520 773__ $$j1$$q25-30$$tProceedings of the International Conference on Artificial Neural Networks (ICANN'99)
000082520 8564_ $$uhttp://publications.idiap.ch/downloads/papers/1999/moerland-density98.pdf$$zURL
000082520 8564_ $$uhttp://publications.idiap.ch/index.php/publications/showcite/moerland-98.2$$zRelated documents
000082520 8564_ $$s190254$$uhttps://infoscience.epfl.ch/record/82520/files/moerland-density98.pdf$$zn/a
000082520 909C0 $$0252189$$pLIDIAP$$xU10381
000082520 909CO $$ooai:infoscience.tind.io:82520$$pconf$$pSTI$$qGLOBAL_SET
000082520 937__ $$aEPFL-CONF-82520
000082520 970__ $$aMoerland-98.2a/LIDIAP
000082520 973__ $$aEPFL$$sPUBLISHED
000082520 980__ $$aCONF