摘要
目的:设计一个慢性乙型肝炎用药推荐系统,对该类住院患者的临床用药进行预测,解决医生主观用药和不规范用药等问题。方法:运用机器学习中的改进惰性多标记分类(multi-label k-nearest neighbor,ML-KNN)算法,对从医院信息系统数据库中抽取的多标记数据集进行训练,并利用Java的相关框架技术在My Eclipse9.0开发平台上设计和实现该慢性乙型肝炎的用药推荐系统。结果:该用药推荐系统能够很好地预测慢性乙型肝炎住院患者的临床用药,提升了住院患者的满意度。结论:基于改进ML-KNN算法的慢性乙型肝炎用药推荐系统对慢性乙型肝炎住院患者的临床用药具有一定的参考意义和实用价值,值得推广使用。
Objective To design and implement a medication recommending system for chronic hepatitis B（CHB） in order to predict CHB inpatient clinical medication. Methods The system was designed and realized with improved ML-KNN algorithm, training the multi-label data set extracted from HIS database, Java framework technology as well as My Eclipse9.0platform. Results The system could predict the clinical medication of the CHB inpatient, and enhanced the CHB inpatient satisfaction greatly. Conclusion The system is of practical value for the the clinical medication of the CHB inpatient, and thus is worthy promoting practically.
出处
《医疗卫生装备》
CAS
2017年第7期48-51,共4页
Chinese Medical Equipment Journal
关键词
慢性乙型肝炎用药推荐系统
多标记分类
算法
JAVA语言
患者满意度
medication recommending system for chronic hepatitis B
multi-label classification
algorithm
Java language
patient satisfaction
