基于深度半非负矩阵分解的医保异常行为检测

医疗保险是一种提供因疾病或伤害而支付的保险福利,检测医保数据和严厉打击欺诈骗保行为是维持医保系统健康发展的必要手段。该文利用机器学习技术,收集医疗保险大数据,通过对原始数据进行降维和聚类分析,对医保数据行为进行异常检测,最终对是否存在欺诈骗保行为作出正确决策。

Medical insurance is a kind of insurance benefits paid for illness or injury. Detecting medical insurance data and cracking down on insurance fraud are necessary means to maintain the healthy development of medical insurance system.In this paper, machine learning technology is used to big data about collect medical insurance, and abnormal detection of medical insurance data behavior is conducted through downgrading and clustering analysis of original data, so as to make correct decisions on whether there is fraud or not.