Research On Sequential Pattern Mining Algorithm In Recommendation Of Hypertensive Drugs

How to find the interesting and valuable knowledge in the massive medical data has become a hot research topic.It is significance that researching the drug treatment of hypertension,found hidden valuable knowledge,scientific and reasonable to provide assistance and recommendations for clinical diagnosis and treatment,avoid medical abnormal behavior.In order to solve the above problems,this paper has done a lot of research on the treatment of hypertension,sequential pattern mining,recommendation algorithm and so on,mainly completed the following work:(1)Aiming at the problem of static assignment of support threshold in traditional sequential pattern mining algorithm,this paper proposes a sequential pattern mining algorithm based on support threshold dynamic selection.According to the relationship between the number of frequent sequences and the support threshold,the relation model of the support threshold with the number of frequent sequences is c onstructed by using the statistical method of curve fitting.The three data sets(Sign Language,BIBLE,and MSNBC)are used to test,and the results show that the algorithm is reasonable and effective.(2)The recommendation algorithm based on rules in the application of antihypertensive drugs recommended,because of only considering the correlation between drugs,regardless of other physical indicators of patients,so the algorithm is la ck of scientific and rational.In this paper,we propose a weighted sequence rule recommendation algorithm for the treatment of hypertension.According to the risk stratification of hypertension,the prescription of drugs was weighted.And the obtained sequence rules with weights are used for recommendation.(3)Based on the electronic medical records of patients with hypertension,based on the above two algorithms,the frequent sequence and recommendation of drug therapy have been studied.The preprocessing data with weight is trained to obtain the parameters of the relational model.Because the results will be used as the knowledge base of recommendation,so the choice of support degree is 0.0013,mining to get frequent sequences with weights.And then the frequent sequence is converted to the sequence rule with weights then used for recommendation.Through the determination of the support threshold and the recommendation based on th e weighted sequence rules,the scientific and rationality of the recommendation of the hypertension drug is improved,and the next step will be the implementation of the recommendation system.