医疗价格异常检测技术研究

打开文本图片集

Abstract：Withtherapidexpansionofthe medicalmarket，abnormalbehaviorssuchas fraudulentdiagnosisandtreatment， excesive fees，and price manipulation have seriouslyundermined marketfairnessand harmed therightsand interestsof patients.Inviewof the technicalproblems of medical price anomalydetection，this paperproposesatechnicalframework based on“datacolection—preprocessing—feature engineering—model detection—result application”，toachieve accurate identification of abnormal behaviors through multi-sourcedata fusionand intellgent modeling.Experiments show that the system achieves a complex abnormal recall rate of 85% in the anti-fraud scenario of the medical insurance fund，and shortensthecase discoveryperiodto25days，providing efective technicalsupportformedicalmarket supervision.InthisstudyGraphNeuralNetwork（GNN）areinnovativelyintroducedintoanomalytraceabilityanalysis， whichsignificantly improves theabilitytoidentifycomplexfraudpatternsandprovidesatheoreticalbasisandtechnical reserve for the modernization of medical price governance.

Keywords：medical prices;anomaly detection;data processing;rules engine;machine learning

0 引言

维护市场秩序的核心要素。（剩余5529字）