How auto-encoders could provide credit assignment in deep networks via target propagation
Y Bengio - arXiv preprint arXiv:1407.7906, 2014 - arxiv.org
Abstract: We propose to exploit {\ em reconstruction} as a layer-local training signal for deep
learning. Reconstructions can be propagated in a form of target propagation playing a role
similar to back-propagation but helping to reduce the reliance on derivatives in order to ...
learning. Reconstructions can be propagated in a form of target propagation playing a role
similar to back-propagation but helping to reduce the reliance on derivatives in order to ...
[PDF][PDF] How Auto-Encoders Could Provide Credit Assignment in Deep Networks via Target Propagation
Y Bengio - arXiv preprint arXiv:1407.7906, 2014 - pdfs.semanticscholar.org
Abstract We propose to exploit reconstruction as a layer-local training signal for deep
learning. Reconstructions can be propagated in a form of target propagation playing a role
similar to back-propagation but helping to reduce the reliance on derivatives in order to ...
learning. Reconstructions can be propagated in a form of target propagation playing a role
similar to back-propagation but helping to reduce the reliance on derivatives in order to ...
How Auto-Encoders Could Provide Credit Assignment in Deep Networks via Target Propagation
Y Bengio - arXiv preprint arXiv:1407.7906, 2014 - adsabs.harvard.edu
Abstract We propose to exploit {\ em reconstruction} as a layer-local training signal for deep
learning. Reconstructions can be propagated in a form of target propagation playing a role
similar to back-propagation but helping to reduce the reliance on derivatives in order to ...
learning. Reconstructions can be propagated in a form of target propagation playing a role
similar to back-propagation but helping to reduce the reliance on derivatives in order to ...