Show simple item record

dc.contributor.authorAn, Senjian
dc.contributor.authorBoussaid, F.
dc.contributor.authorBennamoun, M.
dc.contributor.authorSohel, F.
dc.date.accessioned2018-08-08T04:41:30Z
dc.date.available2018-08-08T04:41:30Z
dc.date.created2018-08-08T03:50:33Z
dc.date.issued2018
dc.identifier.citationAn, S. and Boussaid, F. and Bennamoun, M. and Sohel, F. 2018. Exploiting layerwise convexity of rectifier networks with sign constrained weights. Neural Networks. 105: pp. 419-430.
dc.identifier.urihttp://hdl.handle.net/20.500.11937/69564
dc.identifier.doi10.1016/j.neunet.2018.06.005
dc.description.abstract

© 2018 Elsevier Ltd By introducing sign constraints on the weights, this paper proposes sign constrained rectifier networks (SCRNs), whose training can be solved efficiently by the well known majorization–minimization (MM) algorithms. We prove that the proposed two-hidden-layer SCRNs, which exhibit negative weights in the second hidden layer and negative weights in the output layer, are capable of separating any number of disjoint pattern sets. Furthermore, the proposed two-hidden-layer SCRNs can decompose the patterns of each class into several clusters so that each cluster is convexly separable from all the patterns from the other classes. This provides a means to learn the pattern structures and analyse the discriminant factors between different classes of patterns. Experimental results are provided to show the benefits of sign constraints in improving classification performance and the efficiency of the proposed MM algorithm.

dc.publisherPergamon, Elsevier
dc.titleExploiting layerwise convexity of rectifier networks with sign constrained weights
dc.typeJournal Article
dcterms.source.volume105
dcterms.source.startPage419
dcterms.source.endPage430
dcterms.source.issn0893-6080
dcterms.source.titleNeural Networks
curtin.departmentSchool of Electrical Engineering, Computing and Mathematical Science (EECMS)
curtin.accessStatusFulltext not available


Files in this item

FilesSizeFormatView

There are no files associated with this item.

This item appears in the following Collection(s)

Show simple item record