Component Reduction for Gaussian Mixture Models
Summary :
The mixture modeling framework is widely used in many applications. In this paper, we propose a component reduction technique, that collapses a Gaussian mixture model into a Gaussian mixture with fewer components. The EM (Expectation-Maximization) algorithm is usually used to fit a mixture model to data. Our algorithm is derived by extending mixture model learning using the EM-algorithm. In this extension, a difficulty arises from the fact that some crucial quantities cannot be evaluated analytically. We overcome this difficulty by introducing an effective approximation. The effectiveness of our algorithm is demonstrated by applying it to a simple synthetic component reduction task and a phoneme clustering problem.
- Publication
- IEICE TRANSACTIONS on Information Vol.E91-D No.12 pp.2846-2853
- Publication Date
- 2008/12/01
- Publicized
- Online ISSN
- 1745-1361
- DOI
- 10.1093/ietisy/e91-d.12.2846
- Type of Manuscript
- PAPER
- Category
- Pattern Recognition
Authors
Keyword
Latest Issue
Copyrights notice of machine-translated contents
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Cite this
Copy
Kumiko MAEBASHI, Nobuo SUEMATSU, Akira HAYASHI, "Component Reduction for Gaussian Mixture Models" in IEICE TRANSACTIONS on Information,
vol. E91-D, no. 12, pp. 2846-2853, December 2008, doi: 10.1093/ietisy/e91-d.12.2846.
Abstract: The mixture modeling framework is widely used in many applications. In this paper, we propose a component reduction technique, that collapses a Gaussian mixture model into a Gaussian mixture with fewer components. The EM (Expectation-Maximization) algorithm is usually used to fit a mixture model to data. Our algorithm is derived by extending mixture model learning using the EM-algorithm. In this extension, a difficulty arises from the fact that some crucial quantities cannot be evaluated analytically. We overcome this difficulty by introducing an effective approximation. The effectiveness of our algorithm is demonstrated by applying it to a simple synthetic component reduction task and a phoneme clustering problem.
URL: https://global.ieice.org/en_transactions/information/10.1093/ietisy/e91-d.12.2846/_p
Copy
@ARTICLE{e91-d_12_2846,
author={Kumiko MAEBASHI, Nobuo SUEMATSU, Akira HAYASHI, },
journal={IEICE TRANSACTIONS on Information},
title={Component Reduction for Gaussian Mixture Models},
year={2008},
volume={E91-D},
number={12},
pages={2846-2853},
abstract={The mixture modeling framework is widely used in many applications. In this paper, we propose a component reduction technique, that collapses a Gaussian mixture model into a Gaussian mixture with fewer components. The EM (Expectation-Maximization) algorithm is usually used to fit a mixture model to data. Our algorithm is derived by extending mixture model learning using the EM-algorithm. In this extension, a difficulty arises from the fact that some crucial quantities cannot be evaluated analytically. We overcome this difficulty by introducing an effective approximation. The effectiveness of our algorithm is demonstrated by applying it to a simple synthetic component reduction task and a phoneme clustering problem.},
keywords={},
doi={10.1093/ietisy/e91-d.12.2846},
ISSN={1745-1361},
month={December},}
Copy
TY - JOUR
TI - Component Reduction for Gaussian Mixture Models
T2 - IEICE TRANSACTIONS on Information
SP - 2846
EP - 2853
AU - Kumiko MAEBASHI
AU - Nobuo SUEMATSU
AU - Akira HAYASHI
PY - 2008
DO - 10.1093/ietisy/e91-d.12.2846
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E91-D
IS - 12
JA - IEICE TRANSACTIONS on Information
Y1 - December 2008
AB - The mixture modeling framework is widely used in many applications. In this paper, we propose a component reduction technique, that collapses a Gaussian mixture model into a Gaussian mixture with fewer components. The EM (Expectation-Maximization) algorithm is usually used to fit a mixture model to data. Our algorithm is derived by extending mixture model learning using the EM-algorithm. In this extension, a difficulty arises from the fact that some crucial quantities cannot be evaluated analytically. We overcome this difficulty by introducing an effective approximation. The effectiveness of our algorithm is demonstrated by applying it to a simple synthetic component reduction task and a phoneme clustering problem.
ER -
English
