微型电脑应用2025，Vol.41Issue(3)：15-19,5.
基于改进KNN算法的医院信息系统异常检测系统设计
Design of Anomaly Detection System of Hospital Information System Based on Improved KNN Algorithm
摘要
Abstract
This paper proposes an improved K-nearest neighbor(KNN)algorithm to improve the efficiency of anomaly detection in hospital information system.Each fault is locally diagnosed using KNN to obtain a prior probability.Combined with Gi-uselme distance,the KNN algorithm is improved.The hospital information system is used to analyze the system instability,terminal fault status and other abnormal test data.Through exploring the algorithm performance and system accuracy before and after the improvement,the effectiveness of the proposed improved KNN algorithm is further proved.The research results show that the improved KNN algorithm has good system anomaly detection performance,and has an accuracy rate of 99.38％for anomaly detection in unstable hospital systems.Moreover,the convergence speed of the improved KNN algorithm is 24 times faster than that of the original KNN algorithm.The research achievements can provide a reference for the theoretical and application systems of the key technologies of hospital information systems and the design of anomaly detection systems.关键词
KNN算法/医院信息/系统异常检测/检测准确率Key words
KNN algorithm/hospital information/system anomaly detection/detection accuracy rate分类
信息技术与安全科学引用本文复制引用
王弢,金蕾,赵雪峰..基于改进KNN算法的医院信息系统异常检测系统设计[J].微型电脑应用,2025,41(3):15-19,5.基金项目
教育部产学研合作育人项目(2022023710435) （2022023710435）
浙江省青年科学基金项目(21ZJ190832) （21ZJ190832）
