Number of components and initialization in Gaussian mixture model for pattern recognition

0410563 - UTIA-B 20010032 RIV AT eng C - Konferenční příspěvek (zahraniční konf.)
Paclík, P. - Novovičová, Jana
Number of components and initialization in Gaussian mixture model for pattern recognition.
Wien: Springer, 2001. ISBN 3-211-83651-9. In: Artificial Neural Nets and Genetic Algorithms. Proceedings. - (Kůrková, J.; Neruda, R.; Kárný, M.; Steele, N.), s. 406-409
[International Conference on Artificial Neural Nets and Genetic Algorithms /5./. Prague (CZ), 22.04.2001-25.04.2001]
Grant CEP: GA MŠMT VS96063; GA AV ČR KSK1075601
Výzkumný záměr: AV0Z1075907
Klíčová slova: pattern recognition * Gaussian mixture model * kernel density estimate
Kód oboru RIV: BB - Aplikovaná statistika, operační výzkum

The method for complete mixture initialization based on a product kernel estimate of probability density function is proposed for mixture estimation using EM-algorithm. The mixture components are assumed to correspond to local maxima of optimaly smoothed kernel density estimate. The gradient method is used for local extrema finding. As the last step, agglomerative hiearchical clustering methods merges closest components together. A comparison to scale-space approaches is given on examples.
Trvalý link: http://hdl.handle.net/11104/0130652