Research On Drug Management Of A Hospital Based On Data Mining Technology

With the complexity of diseases and the diversification of drug types,hospital drug inventory,shortage,or oversupply leads to overstock and waste.Accurately grasping of the law of demand for hospital drugs is of great benefit to drug management,so correct prediction of drug dosage is an urgent practical problem for hospital managers to solve.On the other hand,with the popularity of data applications as well as vigorous development of the information management system in hospital,the accumulated amount of data has rapid growth,and it reflects the situation of illness and medication.How to make full use of these data mining and find drug use knowledge and rules in a more scientific and effective way to achieve the purpose of drug management,become the problem the hospital managers have to face,it is also one of research issues of common concern to society.Based on the background above,this paper designed a drug dosage prediction model for scientific drug management combined with time series method.Drug dosage prediction is to extract the law of drug demand from historical data and use this law to predict future drug dosage.For hospitals,this is an important reference for resourcing allocation.In addition,in order to predict the number of patients having certain disease,the law of drug use was mastered from the side,and the prediction model in the number of patients with hypertension was designed by taking the representative hypertension disease as a case which aims to grasp the future drug demand from another perspective.The research content includes the following aspects:The application and research status of data mining technology in the medical field,the theory of data mining technology and the common analysis methods are introduced in detail.On the basis of studying the literature on the research methods of time series prediction,this paper,from the perspective of extracting the potential timing information of data,makes statistical analysis on the historical drug dosage of hospitals and establishes the drug dosage prediction model with the data mining tool--SPSS Modeler,and evaluates the prediction model.Then,statistical analysis was made on the patients' medical treatment,and the demand for drugswas considered from the perspective of the number of patients,aiming to achieve the purpose of scientific management of drugs.SPSS Modeler was also used to establish the prediction model on the number of visits,which was evaluated and analyzed.The predicted results were accurate.Finally,specific decision-making suggestions are given to hospital drug information management.The study in this paper is of practical significance for mastering the drug dosage in hospitals.The hospital can use SPSS Modeler time series modeling node to predict the drug dosage,furthering drug management.