Application Research On Diabetes Patients Information System Based On Data Mining

With the continuous development of intelligent medical health,it has become the focus of medical informationization for data mining from complex medical data to find out the medical laws,which are useful for the treatment and prevention of diseases.As an important branch of data mining technology,association rule mining is a machine learning means to excavate interesting rules from historical data.It can be applied to medical health data mining to find useful medical rules from previous diagnostic and historical treatment records.According to interrelated data,the incidence of diabetes is increasing every year.The complications caused by diabetes can lead to metabolic disorders in patients,long-term medication brings enormous pressure to individuals,families and society.Therefore,by mining and analyzing the disease diagnosis data of the diabetes patient information system,the relevant rules can assist doctors to estimate what complications will occur in diabetes patients,so that medical resources can be reasonably used.In this thesis,the medical diagnosis data of diabetes patient in a tertiary first-class hospital in Nanning of Guangxi were selected.Improved association rule method will be used to explore the correlation of diabetes complications.The mining process will be interfaced and the data mining prototype system for diabetic complications and patient information system will be designed.The following is the majority of the work for this thesis:(1)Data Pretreatment.Creating a diagnosis database of diabetes patients,which diabetes patients are used as data sources.Using Oracle's PL/SQL Developer query tool to filter disease diagnosis data and preprocess the data,and the basic analysis of the data is carried out.(2)Improvement of association rule mining Algorithm Apriori.Taking the Apriori algorithm as the main research focus,a simple example is used to illustrate the bottleneck problem of the algorithm in terms of time complexity and space complexity,as well as the performance problems such as the Apriori algorithm traversing the database several times and the exponential increase of the resulting intermediate candidate sets,and the low utilization rate of I/O.This paper proposes an Avanced_Apriori improved algorithm that can effectively reduce the time and space costs.The basic principle of the Advanced_Apriori algorithm is introduced in detail.At the same time,the implementation process of the improved algorithm is explained through concrete examples.Finally,the effectiveness in improving the efficiency of the original algorithm is illustrated through three sets of experiments.(3)Design and Implementation of Diabetes Patient Information System Based on Data Mining.The paper aims at mining association rules for diabetic complications.The patient information system were designed and implemented based on Advanced Apriori algorithm,and the function of association rules mining and the data preprocessing function were implemented.The system has practical reference value in assisting doctors and patients to prevent and interposediabetic complications.