Affiliations: [a]
School of Electrical Engineering & Automation, Xiamen University of Technology, Xiamen, China
| [b]
Fujian Key Lab of Medical Instrument and Pharmaceutical Technology, Fuzhou University, Fuzhou, China
| [c]
Department of Electrical Engineering and Innovation Center for Biomedical and Healthcare Technology, Yuan Ze University, Chung-Li, Taoyuan, Taiwan
Correspondence:
[*]
Corresponding author. Chih-Min Lin, Department of Electrical Engineering and Innovation Center for Biomedical and Healthcare Technology,
Yuan Ze University, Chung-Li, Taoyuan, Taiwan. Tel.: +886 3 4638800 2202; Fax: +886 3 4639355; E-mail: [email protected].
Abstract: This paper aims to propose a more efficient algorithm for the multi-dimensional classifier design. A novel model of wavelet fuzzy brain emotional learning neural network (WFBELNN) is proposed. This model comprises a wavelet function, a fuzzy inference system and a brain emotional learning neural network. As a result, the learning speed and the classifying accuracy can be effectively improved by the proposed model. The structure of WFBELNN is constructed first, and then the gradient-descent method is used to online tune the parameters of WFBELNN. Finally a medical pattern recognition system is studied to verify that the accurate multi-dimensional pattern recognition can be achieved by using the proposed model. A comparison between the proposed WFBELNN and other models shows that the proposed model can achieve the most accurate classification of the medical pattern recognition and it is also more suitable to deal with the influence of the uncertainties.
