A modified ELM algorithm for single-hidden layer feedforward neural networks with linear nodes

Author

Man, Zhihong ; Lee, Kevin ; Wang, Dianhui ; Cao, Zhenwei ; Miao, Chunyan

Author_Institution

Fac. of Eng. & Ind. Sci., Swinburne Univ. of Technol., Hawthorn, VIC, Australia

fYear

2011

fDate

21-23 June 2011

Firstpage

2524

Lastpage

2529

Abstract

A modified ELM algorithm for a class of single-hidden layer feedforward neural networks (SLFNs) with linear nodes is discussed in this paper. It is seen that the input weights of the SLFN are designed such that the hidden layer performs as a preprocessor for removing the effects of the input disturbance and reducing both the structural and the empirical risks, the output weights are then trained to minimize the output error and further balance and reduce the structural and the empirical risks of the SLFN. The performance of an SLFN-based classifier trained with the proposed scheme is evaluated in the simulation section in support of the proposed scheme.

Keywords

feedforward neural nets; learning (artificial intelligence); pattern classification; SLFN-based classifier; empirical risk reduction; extreme learning machine; input disturbance effect removal; linear nodes; modified ELM algorithm; output error minimization; output weight; single-hidden layer feedforward neural network; structural risk reduction; Algorithm design and analysis; Biological neural networks; Finite impulse response filter; Machine learning; Robustness; Signal to noise ratio; Vectors; extreme learning machine; neural networks; pre-processor; signal classification;

fLanguage

English

Publisher

ieee

Conference_Titel

Industrial Electronics and Applications (ICIEA), 2011 6th IEEE Conference on

Conference_Location

Beijing

ISSN

pending

Print_ISBN

978-1-4244-8754-7

Electronic_ISBN

pending

Type

conf

DOI

10.1109/ICIEA.2011.5976017

Filename

5976017