首页> 外国专利> Deep Neural Network class METHOD AND APPARATUS FOR CLASSIFYING A CLASS TO WHICH A SENTENCE BELONGS BY USING DEEP NEURAL NETWORK

Deep Neural Network class METHOD AND APPARATUS FOR CLASSIFYING A CLASS TO WHICH A SENTENCE BELONGS BY USING DEEP NEURAL NETWORK

机译:深度神经网络类使用深度神经网络对句子分类的分类方法和装置

摘要

A method and apparatus for classifying a class to which a sentence belongs by using deep neural network are provided. According to an embodiment of the present disclosure, a first sentence and a second sentence are respectively learned through a first neural network and a second neural network. Based on whether A first feature vector and a second feature vector generated from the output data of the learning is same as a class to which the first sentence and the second sentence belongs, a contrast loss is obtained. The learning is repeated so as to maximize the contrast loss value. It is possible to improve the classification accuracy of the first sentence.
机译:提供了一种通过使用深度神经网络对句子所属的类别进行分类的方法和设备。根据本公开的实施例,通过第一神经网络和第二神经网络分别学习第一句子和第二句子。基于从学习的输出数据生成的第一特征向量和第二特征向量是否与第一句子和第二句子所属的类相同,获得对比度损失。重复学习以便最大化对比度损失值。可以提高第一句的分类准确性。

著录项

  • 公开/公告号KR20180125905A

    专利类型

  • 公开/公告日2018-11-26

    原文格式PDF

  • 申请/专利权人 SAMSUNG ELECTRONICS CO. LTD.;

    申请/专利号KR20180055651

  • 发明设计人 SONG HEE JUN;KULKARNI NILESH;

    申请日2018-05-15

  • 分类号G06F17/30;G06N3/04;G06N3/08;

  • 国家 KR

  • 入库时间 2022-08-21 11:52:27

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号