基于Word2vec与BP神经网络的病历症状自动分类研究 |
投稿时间:2018-09-04 点此下载全文 |
引用本文:叶辉,卓奕荣,肖志毅,等.基于Word2vec与BP神经网络的病历症状自动分类研究[J].医学信息学杂志,2018,39(11):59-62 |
摘要点击次数: |
全文下载次数: |
|
基金项目:2017国家重点研发计划“云计算和大数据”重点专项(项目编号:2017YFB1002300);2016广东省高水平大学建设青年创新人才项目(项目编号:2016KQNCX024)。 |
|
中文摘要:分别运用Word2vec和BP神经网络算法以及两者混合对300份长病历数据进行症状信息的分类抽取,旨在解决医学病历症状信息自动分类问题,结果表明采用两种方法混合处理能够得到更好的效果。 |
中文关键词:BP神经网络 Word2vec 文本分类 病历症状信息 |
|
Study on the Automatic Classification of Symptoms in Medical Records Based on Word2vec and BP Neural Network |
|
|
Abstract:By making use of Word2vec, BP neural network algorithm and the method mixing the both, respectively, the classification extraction on data of 300 long medical records based on data of symptoms has been implemented, which aims to solve the automatic classification of data of symptoms in medical records. Results show that effect of mixing the two methods in process is better. |
keywords:BP neural network Word2vec text classification symptom information of medical records |
查看全文 查看/发表评论 下载PDF阅读器 |