Authors
Oriol Vinyals, Samy Bengio, Manjunath Kudlur
Publication date
2015/11/19
Journal
arXiv preprint arXiv:1511.06391
Description
Abstract: Sequences have become first class citizens in supervised learning thanks to the
resurgence of recurrent neural networks. Many complex tasks that require mapping from or
to a sequence of observations can now be formulated with the sequence-to-sequence
(seq2seq) framework which employs the chain rule to efficiently represent the joint
probability of sequences. In many cases, however, variable sized inputs and/or outputs
might not be naturally expressed as sequences. For instance, it is not clear how to input a ...
resurgence of recurrent neural networks. Many complex tasks that require mapping from or
to a sequence of observations can now be formulated with the sequence-to-sequence
(seq2seq) framework which employs the chain rule to efficiently represent the joint
probability of sequences. In many cases, however, variable sized inputs and/or outputs
might not be naturally expressed as sequences. For instance, it is not clear how to input a ...
Total citations
Scholar articles
O Vinyals, S Bengio, M Kudlur - arXiv preprint arXiv:1511.06391, 2015
Dates and citation counts are estimated and are determined automatically by a computer program.