Supervised Sequence Labelling with Recurrent Neural Networks
E736828
Supervised Sequence Labelling with Recurrent Neural Networks is a foundational monograph that systematically presents the theory, architectures, and training methods for applying recurrent neural networks to tasks such as speech recognition, handwriting recognition, and other sequence labeling problems.
All labels observed (1)
| Label | Occurrences |
|---|---|
| Supervised Sequence Labelling with Recurrent Neural Networks canonical | 1 |
Statements (44)
| Predicate | Object |
|---|---|
| instanceOf |
book
ⓘ
scientific monograph ⓘ |
| aimsTo | provide a systematic treatment of RNNs for sequence labelling ⓘ |
| author | Alex Graves NERFINISHED ⓘ |
| context | neural network approaches to sequence processing ⓘ |
| covers |
handwriting recognition
ⓘ
labeling unsegmented sequence data ⓘ offline handwriting recognition ⓘ online handwriting recognition ⓘ phoneme recognition ⓘ speech recognition ⓘ |
| emphasizes |
handling variable-length input sequences
ⓘ
learning from unsegmented sequence data ⓘ |
| explains |
LSTM architecture
ⓘ
bidirectional LSTM ⓘ decoding methods for sequence labelling ⓘ recurrent neural network architectures ⓘ regularization for RNNs ⓘ training algorithms for RNNs ⓘ |
| field |
artificial intelligence
ⓘ
handwriting recognition ⓘ machine learning ⓘ pattern recognition ⓘ speech recognition ⓘ |
| focusesOn |
end-to-end training on unsegmented data
ⓘ
supervised sequence learning ⓘ |
| introduces | connectionist temporal classification loss ⓘ |
| isConsidered | foundational work on RNN-based sequence labelling ⓘ |
| language | English ⓘ |
| mainTopic |
backpropagation through time
ⓘ
bidirectional recurrent neural networks ⓘ connectionist temporal classification ⓘ gradient-based training ⓘ long short-term memory ⓘ recurrent neural networks NERFINISHED ⓘ sequence labelling ⓘ sequence modelling ⓘ supervised learning ⓘ |
| provides |
experimental results on sequence labelling tasks
ⓘ
theoretical analysis of RNN training ⓘ |
| targetAudience |
graduate students in computer science
ⓘ
researchers in machine learning ⓘ |
| usedIn |
academic research
ⓘ
graduate-level teaching on neural networks ⓘ |
Referenced by (1)
Full triples — surface form annotated when it differs from this entity's canonical label.