Combining Gaussian Mixture Models

@incollection{citeulike:1813430, abstract = {A Gaussian mixture model (GMM) estimates a probability density function using the expectation-maximization algorithm. However, it may lead to a poor performance or inconsistency. This paper analytically shows that performance of a GMM can be improved in terms of Kullback-Leibler divergence with a committee of GMMs with different initial parameters. Simulations on synthetic datasets demonstrate that a committee of as few as 10 models outperforms a single model.}, author = {Lee, Hyoung-Joo and Cho, Sungzoon }, citeulike-article-id = {1813430}, journal = {Intelligent Data Engineering and Automated Learning – IDEAL 2004}, keywords = {gmm, mixture-models}, pages = {666--671}, posted-at = {2007-10-24 01:52:44}, priority = {2}, title = {Combining Gaussian Mixture Models}, url = {http://www.springerlink.com/content/r95h1caby4g58mne }, year = {2004} }

TY - CHAP ID - citeulike:1813430 L3 - citeulike-article-id:1813430 TI - Combining Gaussian Mixture Models JF - Intelligent Data Engineering and Automated Learning – IDEAL 2004 SP - 666 EP - 671 N2 - A Gaussian mixture model (GMM) estimates a probability density function using the expectation-maximization algorithm. However, it may lead to a poor performance or inconsistency. This paper analytically shows that performance of a GMM can be improved in terms of Kullback-Leibler divergence with a committee of GMMs with different initial parameters. Simulations on synthetic datasets demonstrate that a committee of as few as 10 models outperforms a single model. KW - gmm KW - mixture-models AU - Lee, Hyoung-Joo AU - Cho, Sungzoon PY - 2004/// UR - http://www.springerlink.com/content/r95h1caby4g58mne ER -