引用本文:艾新龑,毛文涛,田梅.基于机器学习技术的在线疾病诊疗方案倾向性识别研究[J].中华医学图书情报杂志,2018,27(7):1-5.
基于机器学习技术的在线疾病诊疗方案倾向性识别研究
Machine learning technique-based tendency cognition of online disease diagnosis and treatment plan
DOI:10.3969/j.issn.1671-3982.2018.07.001
中文关键词:  机器学习  在线医疗  中西医结合  心血管疾病  倾向性识别  文本挖掘
英文关键词:Machine learning  Online medical treatment  Combined Western and traditional Chinese medicine  Cardiovascular disease  Tendency cognition  Text mining
基金项目:国家自然科学基金“基于多任务学习的机械结构小损伤检测方法研究”(U1704158);河南省高校科技创新人才资助计划“基于多任务学习的结构振动微损伤识别方法研究”(15HASTIT022)
作者单位
艾新龑 河南师范大学计算机与信息工程学院河南 新乡 453007 
毛文涛 河南师范大学计算机与信息工程学院河南 新乡 453007 
田梅 新乡医学院管理学院河南 新乡 453003 
摘要点击次数: 399
全文下载次数: 159
中文摘要:
      目的:采用机器学习技术分析和预测在线疾病诊疗方案的倾向性。方法:爬取39疾病百科网中与心血管疾病相关的439条症状数据,通过TF-IDF算法提取症状关键词向量,采用支持向量机、决策树、神经网络建立分类模型,预测不同病症倾向西医或中西医结合的治疗方案。结果:对在线疾病信息的文本分析可挖掘疾病的特征,采用机器学习技术可有效预测对应治疗方案的倾向性,预测精度均达90%以上。结论:机器学习技术可揭示疾病症状和治疗方案之间的内在联系,有助于提高在线疾病咨询的效率,提供有针对性的备选治疗方案。
英文摘要:
      Objective To analyze and predict the tendency for online disease diagnosis and treatment plan using machine learning technique. Methods The data of cardiovascular disease-related 439 symptoms were crawled from The 39 Disease Encyclopedia Web. The keyword vector of symptoms was extracted using TF-IDF algorithm. A classification model was established with support vector machine, decision tree and neural network. The different tendencies of disease to Western medicine or combined Western and traditional Chinese medicine were predicted.Results Text analysis of online disease information could mine the characteristics of different diseases. Machine learning technique could effectively predict the tendency to online disease diagnosis and treatment plan with an accuracy >90%.Conclusion Machine learning technique can display the relationship between disease symptoms and treatment plan, improve the efficiency of online disease consultation, and provide the candidate treatment plan for different diseases.
查看全文  查看/发表评论  下载PDF阅读器
关闭
手机扫一扫看