A modified extreme learning machine with sigmoidal activation functions

Zhixiang X. Chen; Houying Y. Zhu; Yuguang G. Wang

doi:10.1007/s00521-012-0860-2

A modified extreme learning machine with sigmoidal activation functions

Zhixiang X. Chen, Houying Y. Zhu, Yuguang G. Wang

Research output: Contribution to journal › Article › peer-review

42 Citations (Scopus)

Abstract

This paper proposes a modified ELM algorithm that properly selects the input weights and biases before training the output weights of single-hidden layer feedforward neural networks with sigmoidal activation function and proves mathematically the hidden layer output matrix maintains full column rank. The modified ELM avoids the randomness compared with the ELM. The experimental results of both regression and classification problems show good performance of the modified ELM algorithm.

Original language	English
Pages (from-to)	541–550
Number of pages	10
Journal	Neural Computing and Applications
Volume	22
Issue number	3-4
DOIs	https://doi.org/10.1007/s00521-012-0860-2
Publication status	Published - Mar 2013
Externally published	Yes

Keywords

Feedforward neural networks
Extreme learning machine
Moore–Penrose generalized inverse

Access to Document

10.1007/s00521-012-0860-2

Cite this

@article{3244f6334a8247f985b5f03617fd937b,

title = "A modified extreme learning machine with sigmoidal activation functions",

abstract = "This paper proposes a modified ELM algorithm that properly selects the input weights and biases before training the output weights of single-hidden layer feedforward neural networks with sigmoidal activation function and proves mathematically the hidden layer output matrix maintains full column rank. The modified ELM avoids the randomness compared with the ELM. The experimental results of both regression and classification problems show good performance of the modified ELM algorithm.",

keywords = "Feedforward neural networks, Extreme learning machine, Moore–Penrose generalized inverse",

author = "Chen, {Zhixiang X.} and Zhu, {Houying Y.} and Wang, {Yuguang G.}",

year = "2013",

month = mar,

doi = "10.1007/s00521-012-0860-2",

language = "English",

volume = "22",

pages = "541–550",

journal = "Neural Computing and Applications",

issn = "0941-0643",

publisher = "Springer, Springer Nature",

number = "3-4",

}

TY - JOUR

T1 - A modified extreme learning machine with sigmoidal activation functions

AU - Chen, Zhixiang X.

AU - Zhu, Houying Y.

AU - Wang, Yuguang G.

PY - 2013/3

Y1 - 2013/3

N2 - This paper proposes a modified ELM algorithm that properly selects the input weights and biases before training the output weights of single-hidden layer feedforward neural networks with sigmoidal activation function and proves mathematically the hidden layer output matrix maintains full column rank. The modified ELM avoids the randomness compared with the ELM. The experimental results of both regression and classification problems show good performance of the modified ELM algorithm.

AB - This paper proposes a modified ELM algorithm that properly selects the input weights and biases before training the output weights of single-hidden layer feedforward neural networks with sigmoidal activation function and proves mathematically the hidden layer output matrix maintains full column rank. The modified ELM avoids the randomness compared with the ELM. The experimental results of both regression and classification problems show good performance of the modified ELM algorithm.

KW - Feedforward neural networks

KW - Extreme learning machine

KW - Moore–Penrose generalized inverse

UR - http://www.scopus.com/inward/record.url?scp=84874020734&partnerID=8YFLogxK

U2 - 10.1007/s00521-012-0860-2

DO - 10.1007/s00521-012-0860-2

M3 - Article

SN - 0941-0643

VL - 22

SP - 541

EP - 550

JO - Neural Computing and Applications

JF - Neural Computing and Applications

IS - 3-4

ER -

A modified extreme learning machine with sigmoidal activation functions

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this