Open AccessAn effective classifier based on convolutional neural network and regularized extreme learning machine
Author(s) -
Chuangxin He,
Hong Kang,
Yao Tong,
Xiao Rui Li
Publication year - 2019
Publication title -
mathematical biosciences and engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.451
H-Index - 45
eISSN - 1551-0018
pISSN - 1547-1063
DOI - 10.3934/mbe.2019420
Subject(s) - convolutional neural network , extreme learning machine , computer science , artificial intelligence , classifier (uml) , pattern recognition (psychology) , deep learning , machine learning , artificial neural network
An effective classifier combining convolutional neural network and regularized extreme learning machine (called as CNN-RELM) is presented in this paper. Firstly, CNN-RELM trains the convolutional neural network (CNN) using the gradient descent method until the learning target accuracy reaches. Then the fully connected layer of CNN is replaced by regularized extreme learning machine (RELM) optimized by genetic algorithm and the rest layers of the CNN remain unchanged. The experiments on different face databases are given to evaluate the performance of CNN-RELM. The experimental results show that CNN-RELM is a feasible classifier and it outperforms CNN and RELM. Due to the uniting of CNN and RELM, CNN-RELM have the advantages of CNN and RELM and it is easier to learn and faster in testing.