informИнформатикаInformatics1816-03012617-6963UIIP NASBinform-276Research ArticleОБРАБОТКА СИГНАЛОВ, ИЗОБРАЖЕНИЙ, РЕЧИ, ТЕКСТА И РАСПОЗНАВАНИЕ ОБРАЗОВSIGNAL, IMAGE, SPEECH, TEXT PROCESSING AND PATTERN RECOGNITIONАЛГОРИТМ ОПТИМИЗАЦИИ ВХОДНЫХ ПРИЗНАКОВ И ПАРАМЕТРОВ КЛАССИФИКАТОРА ЭМОЦИОНАЛЬНОГО СОСТОЯНИЯ ЧЕЛОВЕКА ПО ЕГО РЕЧИПекарьД. В.ТихоненкоС. Г.ЧернявскийА. Ф.СадовВ. С.Белорусский государственный университетBelarus20122602201803(35)2634Copyright © Пекарь Д.В., Тихоненко С.Г., Чернявский А.Ф., Садов В.С., 20182018Пекарь Д.В., Тихоненко С.Г., Чернявский А.Ф., Садов В.С.Пекарь Д.В., Тихоненко С.Г., Чернявский А.Ф., Садов В.С.Данная работа распространяется под лицензией Creative Commons Attribution 4.0.This work is licensed under a Creative Commons Attribution 4.0 License.https://inf.grid.by/jour/article/view/276

Рассматривается алгоритм оптимизации входного вектора признаков и парамет-ров классификатора эмоционального состояния человека, основанный на использованиистохастического метода роя частиц. Предлагаются пути его совершенствования с цельюповышения скорости сходимости алгоритма. Доказывается улучшение точности класси-фикации эмоционального состояния диктора по его речи при использовании оптимизиро-ванного вектора признаков и найденных параметров классификатора.

ReferencesWilliams, C.E. Emotions and Speech: Some Acoustical Correlates / C.E. Williams // JournalWilliams, C.E. Emotions and Speech: Some Acoustical Correlates / C.E. Williams // Journalof the Acoustical Society of America. – 1972. – Vol. 52, № 4. – P. 1238–1250.of the Acoustical Society of America. – 1972. – Vol. 52, № 4. – P. 1238–1250.Banse, R. Acoustic profiles in vocal emotion expression / R. Banse // Journal of personalityBanse, R. Acoustic profiles in vocal emotion expression / R. Banse // Journal of personalityand social psychology. – 1996. – Vol. 70, № 3. – P. 614–636.and social psychology. – 1996. – Vol. 70, № 3. – P. 614–636.Eyben, F. OpenEAR – Introducing the Munich Open-Source Emotion and Affect / F. Eyben // Proc. ACII. – Amsterdam, Netherlands, 2009. – P. 576–581.Eyben, F. OpenEAR – Introducing the Munich Open-Source Emotion and Affect / F. Eyben // Proc. ACII. – Amsterdam, Netherlands, 2009. – P. 576–581.Schuller, B. The interspeech 2009 emotion / B. Schuller // Interspeech. – Brighton, UK, 2009.Schuller, B. The interspeech 2009 emotion / B. Schuller // Interspeech. – Brighton, UK, 2009.Eyben, F. OpenSMILE : the Munich open Speech and Music Interpretation by Large SpaceEyben, F. OpenSMILE : the Munich open Speech and Music Interpretation by Large SpaceExtraction toolkit / F. Eyben, M. Woellmer, B. Schuller. – 2010. – Mode of access :Extraction toolkit / F. Eyben, M. Woellmer, B. Schuller. – 2010. – Mode of access :http://sourceforge.net/projects/opensmile/files/openSMILE_book_1.0.1.pdf. – Date of access : 20.01.2012.http://sourceforge.net/projects/opensmile/files/openSMILE_book_1.0.1.pdf. – Date of access : 20.01.2012.Berlin Database of Emotional Speech [Electronic resource]. – Technical University Berlin,Berlin Database of Emotional Speech [Electronic resource]. – Technical University Berlin,– Mode of access : http://pascal.kgw.tu-berlin.de/emodb/index-1024.html. – Date of access : 20.01.2012.– Mode of access : http://pascal.kgw.tu-berlin.de/emodb/index-1024.html. – Date of access : 20.01.2012.Holmes, G. WEKA: a machine learning / G. Holmes // Intelligent Information Systems. –Holmes, G. WEKA: a machine learning / G. Holmes // Intelligent Information Systems. –– № 29. – P. 357–361.– № 29. – P. 357–361.John G.H. Estimating Continuous Distributions in Bayesian Classifiers / G.H. John // Proc. of the Eleventh Conf. on Uncertainty in Artificial Intelligence. – San Francisco, USA, 1995. – Vol. 1. – P. 338–345.John G.H. Estimating Continuous Distributions in Bayesian Classifiers / G.H. John // Proc. of the Eleventh Conf. on Uncertainty in Artificial Intelligence. – San Francisco, USA, 1995. – Vol. 1. – P. 338–345.LIBLINEAR – A Library for Large Linear Classification // Rong-En Fan [et al.]. – 2008. –LIBLINEAR – A Library for Large Linear Classification // Rong-En Fan [et al.]. – 2008. –Mode of access : http://www.csie.ntu.edu.tw/~cjlin/liblinear/. – Date of access : 20.01.2012.Mode of access : http://www.csie.ntu.edu.tw/~cjlin/liblinear/. – Date of access : 20.01.2012.Quinlan, R.J. C4.5: Programs for Machine Learning / R.J. Quinlan // Machine Learning. –Quinlan, R.J. C4.5: Programs for Machine Learning / R.J. Quinlan // Machine Learning. –– Vol. 16, № 3. – P. 235–240.– Vol. 16, № 3. – P. 235–240.Aha, D.W. Instance-Based Learning Algorithms / D.W. Aha // Machine Learning. – Hingham, USA, 1991. – Vol. 6, № 1. – P. 37–66.Aha, D.W. Instance-Based Learning Algorithms / D.W. Aha // Machine Learning. – Hingham, USA, 1991. – Vol. 6, № 1. – P. 37–66.Kennedy, J. Particle Swarm Optimization / J. Kennedy // Proc. of IEEE Intern. Conf. onKennedy, J. Particle Swarm Optimization / J. Kennedy // Proc. of IEEE Intern. Conf. onNeural Networks. – Washington, USA, 1995. – Vol. 4. – P. 1942–1948.Neural Networks. – Washington, USA, 1995. – Vol. 4. – P. 1942–1948.Sun, J. Particle Swarm Optimization with particles having Quantum Behavior / J. Sun //Sun, J. Particle Swarm Optimization with particles having Quantum Behavior / J. Sun //Proc. of Congress on Evolutionary Computation. – Portland, USA, 2004. – P. 325–331.Proc. of Congress on Evolutionary Computation. – Portland, USA, 2004. – P. 325–331.Gao, H. A Simple Quantum-inspired Particle swarm Optimization and its Application /Gao, H. A Simple Quantum-inspired Particle swarm Optimization and its Application /H. Gao // Information Technology Journal. – 2011. – № 10. – P. 2315–2321.H. Gao // Information Technology Journal. – 2011. – № 10. – P. 2315–2321.Maolong, X. An improved quantum-behaved particle swarm optimization algorithm withMaolong, X. An improved quantum-behaved particle swarm optimization algorithm withweighted mean best position / X. Maolong // Applied Mathematics and Computation. – 2008. – № 205. – P. 751–759.weighted mean best position / X. Maolong // Applied Mathematics and Computation. – 2008. – № 205. – P. 751–759.

The authors declare that there are no conflicts of interest present.