Welcome to the InfoSci Platform
Broad Autoencoder Features Learning for Classification Problem

Broad Autoencoder Features Learning for Classification Problem

Ting Wang, Wing W. Y. Ng, Wendi Li, Sam Kwong
Copyright: © 2021 |Volume: 15 |Issue: 4 |Pages: 15
ISSN: 1557-3958|EISSN: 1557-3966|EISBN13: 9781799859857|DOI: 10.4018/IJCINI.20211001.oa23
Cite Article Cite Article

MLA

Wang, Ting, et al. "Broad Autoencoder Features Learning for Classification Problem." IJCINI vol.15, no.4 2021: pp.1-15. http://doi.org/10.4018/IJCINI.20211001.oa23

APA

Wang, T., Ng, W. W., Li, W., & Kwong, S. (2021). Broad Autoencoder Features Learning for Classification Problem. International Journal of Cognitive Informatics and Natural Intelligence (IJCINI), 15(4), 1-15. http://doi.org/10.4018/IJCINI.20211001.oa23

Chicago

Wang, Ting, et al. "Broad Autoencoder Features Learning for Classification Problem," International Journal of Cognitive Informatics and Natural Intelligence (IJCINI) 15, no.4: 1-15. http://doi.org/10.4018/IJCINI.20211001.oa23

Export Reference

Mendeley
Favorite Full-Issue Download

Abstract

Activation functions such as Tanh and Sigmoid functions are widely used in Deep Neural Networks (DNNs) and pattern classification problems. To take advantages of different activation functions, the Broad Autoencoder Features (BAF) is proposed in this work. The BAF consists of four parallel-connected Stacked Autoencoders (SAEs) and each of them uses a different activation function, including Sigmoid, Tanh, ReLU, and Softplus. The final learned features can merge such features by various nonlinear mappings from original input features with such a broad setting. This helps to excavate more information from the original input features. Experimental results show that the BAF yields better-learned features and classification performances.