Deep Learning of Part-Based Representation of Data Using Sparse Autoencoders With Nonnegativity Constraints

We demonstrate a new deep learning autoencoder network, trained by a nonnegativity constraint algorithm (nonnegativity-constrained autoencoder), that learns features that show part-based representation of data. The learning algorithm is based on constraining negative weights. The performance of the...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transaction on neural networks and learning systems Vol. 27; no. 12; pp. 2486 - 2498
Main Authors: Hosseini-Asl, Ehsan, Zurada, Jacek M., Nasraoui, Olfa
Format: Journal Article
Language:English
Published: United States IEEE 01.12.2016
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects:
Algorithms
Artificial neural networks
Autoencoder
Constraints
Cost function
Datasets
deep architectures
Deep learning
Encoding
Feature extraction
feature learning
Image reconstruction
Machine learning
Neural networks
nonnegativity constraints
part-based representation
Performance prediction
Representations
Training
ISSN:2162-237X, 2162-2388, 2162-2388
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We demonstrate a new deep learning autoencoder network, trained by a nonnegativity constraint algorithm (nonnegativity-constrained autoencoder), that learns features that show part-based representation of data. The learning algorithm is based on constraining negative weights. The performance of the algorithm is assessed based on decomposing data into parts and its prediction performance is tested on three standard image data sets and one text data set. The results indicate that the nonnegativity constraint forces the autoencoder to learn features that amount to a part-based representation of data, while improving sparsity and reconstruction quality in comparison with the traditional sparse autoencoder and nonnegative matrix factorization. It is also shown that this newly acquired representation improves the prediction performance of a deep neural network.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:2162-237X
2162-2388
2162-2388
DOI:10.1109/TNNLS.2015.2479223