|Title:||Learning long-term dependencies in segmented-memory recurrent neural networks with backpropagation of error|
|Authors :||Glüge, Stefan|
|Published in :||Neurocomputing|
|Publisher / Ed. Institution :||Elsevier|
|License (according to publishing contract) :||Licence according to publishing contract|
|Type of review:||Peer review (publication)|
|Subject (DDC) :||004: Computer science|
|Abstract:||In general, recurrent neural networks have difficulties in learning long-term dependencies. The segmented-memory recurrent neural network (SMRNN) architecture together with the extended real-time recurrent learning (eRTRL) algorithm was proposed to circumvent this problem. Due to its computational complexity eRTRL becomes impractical with increasing network size. Therefore, we introduce the less complex extended backpropagation through time (eBPTT) for SMRNN together with a layer-local unsupervised pre-training procedure. A comparison on the information latching problem showed that eRTRL is better able to handle the latching of information over longer periods of time, even though eBPTT guaranteed a better generalisation when training was successful. Further, pre-training significantly improved the ability to learn long-term dependencies with eBPTT. Therefore, the proposed eBPTT algorithm is suited for tasks that require big networks where eRTRL is impractical. The pre-training procedure itself is independent of the supervised learning algorithm and can improve learning in SMRNN in general.|
|Departement:||Life Sciences and Facility Management|
|Organisational Unit:||Institute of Applied Simulation (IAS)|
|Publication type:||Article in scientific journal|
|Appears in Collections:||Publikationen Life Sciences und Facility Management|
Files in This Item:
There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.