Biblioteca Digital do Instituto Politécnico de Bragança   Instituto Politécnico de Bragança

Biblioteca Digital do IPB >
Escola Superior de Tecnologia e Gestão >
Electrotecnia >
DE - Artigos em Proceedings Não Indexados ao ISI >

Please use this identifier to cite or link to this item: http://hdl.handle.net/10198/4468

Título: Training hidden markov models with the taguchi method
Autor: Coelho, João
Cunha, José Boaventura
Oliveira, Paulo
Palavras-chave: Hidden Markov models
Taguchi method
Optimization problems
Issue Date: 2010
Citação: Coelho, João; Cunha, José; Oliveira, Paulo (2010) - Training hidden markov models with the taguchi method. In 9th Portuguese Conference on Automatic Control.
Resumo: In some control systems structures, like predictive control, mathematical models for the control process must be derived. Those models can be obtained by a broad class of methods like parametric models applied to experimental data. In this context, and for systems with multiple operation regimes, the Hidden Markov model, due to its properties, is a convincing choice. However the parameter estimation of this type of models involves the optimization of a non-convex cost function. So the Baum-Welch method only can find sub-optimal parameters. This article shows that the use of the Taguchi method minimizes the training algorithm sensibility local minima.
Arbitragem científica: yes
URI: http://hdl.handle.net/10198/4468
Appears in Collections:DE - Artigos em Proceedings Não Indexados ao ISI

Files in This Item:

File Description SizeFormat
34_FINAL.PDF152 kBAdobe PDFView/Open
Restrict Access. You can request a copy!
Training Hidden - resumo.PDF17,28 kBAdobe PDFView/Open

Statistics
FacebookTwitterDeliciousLinkedInDiggGoogle BookmarksMySpaceOrkut
Formato BibTex mendeley Endnote Logotipo do DeGóis 

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

 


  © Instituto Politécnico de Bragança - Biblioteca Digital - Feedback - Statistics
  Estamos no RCAAP Governo Português separator Ministério da Educação e Ciência   Fundação para a Ciência e a Tecnologia

Financiado por:

POS_C UE