Open Access Repository

Neural Transplant Surgery: An Approach to Pre-training Recurrent Networks


Downloads per month over past year

Vamplew, P and Adams, A 1994 , 'Neural Transplant Surgery: An Approach to Pre-training Recurrent Networks', paper presented at the Fifth Australian Conference on Neural Networks, February 1994, University of Queensland.

transplant-acnn...pdf | Download (149kB)
Available under University of Tasmania Standard License.

| Preview


Partially-recurrent networks have advantages over strictly feed-forward networks for certain spatiotemporal pattern classification or prediction tasks. However networks involving recurrent links are generally more difficult to train than their non-recurrent counterparts. In this paper we demonstrate that the costs of training a recurrent network can be greatly reduced by initialising the network prior to training with weights 'transplanted' from a non-recurrent architecture.

Item Type: Conference or Workshop Item (Paper)
Authors/Creators:Vamplew, P and Adams, A
Keywords: recurrent neural networks, pre-training, weight initialisation
Item Statistics: View statistics for this item

Actions (login required)

Item Control Page Item Control Page