|
gptkbp:instanceOf
|
gptkb:convolutional_neural_network
|
|
gptkbp:activatedBy
|
sigmoid
tanh
|
|
gptkbp:advantage
|
faster training than LSTM
may be less expressive than LSTM
simpler architecture than LSTM
|
|
gptkbp:category
|
gptkb:artificial_intelligence
gptkb:machine_learning
deep learning
|
|
gptkbp:contrastsWith
|
gptkb:LSTM
|
|
gptkbp:excludes
|
separate memory cell
|
|
gptkbp:fullName
|
Gated Recurrent Units
|
|
gptkbp:hasCellState
|
hidden state
|
|
gptkbp:hasComponent
|
reset gate
update gate
|
|
gptkbp:hasFewerParametersThan
|
gptkb:LSTM
|
|
gptkbp:introduced
|
gptkb:Yoshua_Bengio
gptkb:Kyunghyun_Cho
gptkb:Dzmitry_Bahdanau
Bart van Merriënboer
|
|
gptkbp:introducedIn
|
2014
|
|
gptkbp:publishedIn
|
arXiv:1406.1078
|
|
gptkbp:relatedTo
|
gptkb:LSTM
|
|
gptkbp:usedFor
|
natural language processing
time series prediction
sequence modeling
|
|
gptkbp:usedIn
|
gptkb:TensorFlow
gptkb:Chainer
gptkb:Keras
gptkb:MXNet
gptkb:Apache_MXNet
gptkb:CNTK
gptkb:Theano
gptkb:Microsoft_Cognitive_Toolkit
gptkb:PyTorch
machine translation
robotics
speech recognition
bioinformatics
financial forecasting
speech synthesis
recurrent neural networks
handwriting recognition
text generation
anomaly detection
video analysis
music modeling
|
|
gptkbp:bfsParent
|
gptkb:Transformer_models
|
|
gptkbp:bfsLayer
|
7
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
GRUs
|