公开(公告)号：US20160210551A1

公开(公告)日：2016-07-21

申请号：US14992129

申请日：2016-01-11

Applicant: Samsung Electronics Co., Ltd.

Inventor： Hodong LEE ,  Hoshik LEE ,  Heeyoul CHOI ,  Yunhong MIN ,  Sang Hyun YOO ,  Yeha LEE ,  Jihyun LEE ,  YoungSang CHOI

IPC: G06N3/08 ,  G10L15/16 ,  G06F17/27 ,  G06N3/04

CPC classification number: G06N3/08 ,  G06F17/27 ,  G06N3/04 ,  G06N3/0445 ,  G06N3/084 ,  G10L15/005 ,  G10L15/16 ,  G10L15/19

Abstract: A method and apparatus for training a language model, include generating a first training feature vector sequence and a second training feature vector sequence from training data. The method is configured to perform forward estimation of a neural network based on the first training feature vector sequence, and perform backward estimation of the neural network based on the second training feature vector sequence. The method is further configured to train a language model based on a result of the forward estimation and a result of the backward estimation.

Abstract translation: 一种用于训练语言模型的方法和装置，包括从训练数据生成第一训练特征向量序列和第二训练特征向量序列。 该方法被配置为基于第一训练特征向量序列执行神经网络的前向估计，并且基于第二训练特征向量序列执行神经网络的向后估计。 该方法还被配置为基于正向估计的结果和后向估计的结果来训练语言模型。

公开(公告)号：US20160026913A1

公开(公告)日：2016-01-28

申请号：US14636684

申请日：2015-03-03

Applicant: Samsung Electronics Co., Ltd.

Inventor： Taesup MOON ,  Yeha LEE ,  Heeyoul CHOI

IPC: G06N3/08 ,  G06N3/04

CPC classification number: G06N3/08 ,  G06N3/04 ,  G06N3/0445 ,  G06N3/082

Abstract: A neural network training method based on training data, includes receiving training data including sequential data, and selecting a reference hidden node from hidden nodes in a neural network. The method further includes training the neural network based on remaining hidden nodes obtained by excluding the reference hidden node from the hidden nodes, and based on the training data, the remaining hidden nodes being connected with hidden nodes in a different time interval, and a connection between the reference hidden node and the hidden nodes in the different time interval being ignored.

Abstract translation: 一种基于训练数据的神经网络训练方法，包括接收包括顺序数据的训练数据，以及从神经网络中的隐藏节点中选择参考隐藏节点。 该方法还包括基于通过从隐藏节点排除参考隐藏节点获得的剩余隐藏节点来训练神经网络，并且基于训练数据，剩余的隐藏节点与不同时间间隔中的隐藏节点连接，以及连接 在不同时间间隔内的引用隐藏节点和隐藏节点之间被忽略。