Multiple parallel hidden layers and other improvements to recurrent neural network language modeling.
Diamantino CaseiroAndrej LjoljePublished in: ICASSP (2013)
Keyphrases
- recurrent neural networks
- language modeling
- hidden layer
- language model
- feedforward neural networks
- neural network
- feed forward
- information retrieval
- feed forward neural networks
- learning rate
- retrieval model
- back propagation
- activation function
- artificial neural networks
- probabilistic model
- query expansion
- neural nets
- n gram
- radial basis function
- multilayer perceptron
- rbf neural network
- data mining
- document retrieval
- text classification
- co occurrence
- knowledge discovery
- high dimensional
- similarity measure
- clustering algorithm
- learning algorithm