DocumentCode :
510074
Title :
Modeling of Fermentation Process Based on MOACO and Epsilon-SVM
Author :
Zhu, Jian-ping ; Zhou, Lin-cheng ; Liu, Chun-bo
Author_Institution :
Jiangsu Coll. of Inf. Technol., Wuxi, China
Volume :
2
fYear :
2009
fDate :
7-8 Nov. 2009
Firstpage :
234
Lastpage :
239
Abstract :
To establish suitable models to describe the behavior of biochemistry systems, a new modeling method was introduced, combining multiple objective ant colony optimization(MOACO) with the dynamic Epsilon-SVM. The hyper-parameters of Epsilon-SVM were automatically decided by using multiple objective ant colony optimization(MOACO). Each training sample used different error. The model for penicillin production´s prediction was developed using the method with data collected from real plant in Matlab7.0. The model possesses strong capability of fitting and generalization. Experiments also show that the dynamic Epsilon-SVM is superior to the standard SVM modeling method. MOACO is very feasible and efficient too.
Keywords :
biochemistry; fermentation; mathematics computing; optimisation; pharmaceutical technology; support vector machines; Epsilon-SVM; MOACO; Matlab7.0; biochemistry systems; fermentation process; multiple objective ant colony optimization; Ant colony optimization; Artificial intelligence; Artificial neural networks; Biosensors; Computational intelligence; Computer languages; Mathematical model; Risk management; Support vector machines; Vehicle dynamics; Epsilon-SVM; modeling; multiple objective ant colony optimization (MOACO); support vector machine;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Artificial Intelligence and Computational Intelligence, 2009. AICI '09. International Conference on
Conference_Location :
Shanghai
Print_ISBN :
978-1-4244-3835-8
Electronic_ISBN :
978-0-7695-3816-7
Type :
conf
DOI :
10.1109/AICI.2009.124
Filename :
5375955
Link To Document :
بازگشت