|Appears in Collections:||Computing Science and Mathematics Journal Articles|
|Peer Review Status:||Refereed|
|Title:||Improve deep learning with unsupervised objective|
|Citation:||Zhang S, Huang K, Zhang R & Hussain A (2017) Improve deep learning with unsupervised objective. In: Liu D, Xie S, Li Y, Zhao D & El-Alfy E (eds.) Neural Information Processing. ICONIP 2017. Lecture Notes in Computer Science, 10634. 24th International Conference On Neural Information Processing: ICONIP 2017, Guangzhou, China, 14.11.2017-18.11.2017. Cham, Switzerland: Springer, pp. 720-728. https://link.springer.com/chapter/10.1007/978-3-319-70087-8_74; https://doi.org/10.1007/978-3-319-70087-8_74|
|Series/Report no.:||Lecture Notes in Computer Science, 10634|
|Abstract:||We propose a novel approach capable of embedding the unsupervised objective into hidden layers of the deep neural network (DNN) for preserving important unsupervised information. To this end, we exploit a very simple yet effective unsupervised method, i.e. principal component analysis (PCA), to generate the unsupervised “label" for the latent layers of DNN. Each latent layer of DNN can then be supervised not just by the class label, but also by the unsupervised “label" so that the intrinsic structure information of data can be learned and embedded. Compared with traditional methods which combine supervised and unsupervised learning, our proposed model avoids the needs for layer-wise pre-training and complicated model learning e.g. in deep autoencoder. We show that the resulting model achieves state-of-the-art performance in both face and handwriting data simply with learning of unsupervised “labels".|
|Rights:||Published in Liu D, Xie S, Li Y, Zhao D, El-Alfy E-SM (ed.) Neural Information Processing. ICONIP 2017, Cham, Switzerland: Springer. Part of the Lecture Notes in Computer Science book series (LNCS, volume 10634). The final publication is available at Springer via http://dx.doi.org/10.1007/978-3-319-70087-8_74|
|ShufeiZhang17.pdf||Fulltext - Accepted Version||797.52 kB||Adobe PDF||View/Open|
This item is protected by original copyright
Items in the Repository are protected by copyright, with all rights reserved, unless otherwise indicated.
If you believe that any material held in STORRE infringes copyright, please contact email@example.com providing details and we will remove the Work from public display in STORRE and investigate your claim.