X93-1009 itself very nicely to conventional neural network training algorithms . One of those algorithms
P14-2023 n-best lists with the new weights . Neural Network Training . All neural network models are
P14-1129 of -- 15x . Table 1 shows the neural network training results with various values of
D14-1003 Huck et al. , 2013 ) . For the neural network training , we selected a subset of 9M
W15-3034 using 17 dense fea - tures . The neural network training is performed using the same data
D14-1003 using 17 dense features . The neural network training was performed using a selection
D15-1040 representation learning as part of a neural network training . The underlying hypothesis for
W04-0305 d1 , ... , di − 1 ) . The neural network training methods we use try to find representations
P04-1013 , dm ) ) , respectively . The neural network training methods we use try to find representations
D15-1029 Segmenter with default settings . 3.2 Neural Network Training Training was performed with an
E03-1002 which are induced as part of the neural network training process . These induced features
P14-1013 . There are two phases in our neural network training process : pre-training and fine-tuning
P14-1013 machine translation task . In neural network training , a large number of monolingual
P13-1017 neural net - works . Besides that , neural network training also involves some hyperparameters
P14-1129 are given in Section 6.5 . 2.2 Neural Network Training The training procedure is identical
P05-1023 finite number of these parameters . Neural network training is applied to determine the values
P05-1023 ( d1 , ... , di − 1 ) . Neural network training tries to find such a history
P04-1013 This paper has also proposed a neural network training method which optimizes a discriminative
W13-4707 propagation method is commonly used in neural network training ( V. J. Hodge and J. Austin 2003
D15-1214 inspired by " dropout " as is used in neural network training , where various connections between
hide detail