Please use this identifier to cite or link to this item: http://hdl.handle.net/10397/32651
Title: Deep learning regularized fisher mappings
Authors: Wong, WK 
Sun, M
Keywords: Deep learning architecture
feature extraction
Fisher criterion
regularization
Issue Date: 2011
Publisher: Institute of Electrical and Electronics Engineers
Source: IEEE transactions on neural networks, 2011, v. 22, no. 10, 5982410, p. 1668-1675 How to cite?
Journal: IEEE transactions on neural networks 
Abstract: For classification tasks, it is always desirable to extract features that are most effective for preserving class separability. In this brief, we propose a new feature extraction method called regularized deep Fisher mapping (RDFM), which learns an explicit mapping from the sample space to the feature space using a deep neural network to enhance the separability of features according to the Fisher criterion. Compared to kernel methods, the deep neural network is a deep and nonlocal learning architecture, and therefore exhibits more powerful ability to learn the nature of highly variable datasets from fewer samples. To eliminate the side effects of overfitting brought about by the large capacity of powerful learners, regularizers are applied in the learning procedure of RDFM. RDFM is evaluated in various types of datasets, and the results reveal that it is necessary to apply unsupervised regularization in the fine-tuning phase of deep learning. Thus, for very flexible models, the optimal Fisher feature extractor may be a balance between discriminative ability and descriptive ability.
URI: http://hdl.handle.net/10397/32651
ISSN: 1045-9227
DOI: 10.1109/TNN.2011.2162429
Appears in Collections:Journal/Magazine Article

Access
View full-text via PolyU eLinks SFX Query
Show full item record

SCOPUSTM   
Citations

18
Last Week
1
Last month
0
Citations as of Oct 19, 2017

WEB OF SCIENCETM
Citations

9
Last Week
0
Last month
0
Citations as of Oct 14, 2017

Page view(s)

47
Last Week
5
Last month
Checked on Oct 16, 2017

Google ScholarTM

Check

Altmetric



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.