Research And Improvement Of Apriori Algorithm For Medical Cloud Data Based On Hadoop

Science and technology in the medical field have developed rapidly in recent years.The growing gap between health care costs and results is one of the most important issues.Many developed countries are making many efforts to fill this gap.Therefore,there is an urgent need to improve medical quality,increase data availability and analysis capabilities,which is the driving force in the era of medical big data.In the face of the relatively large scale of medical data,the traditional data mining method seems to be stretched,and a better data mining algorithm has been improved from the traditional data mining method.At the same time,the rise of cloud computing technology has also brought new options for big data processing.This article summarizes the relevant knowledge of data mining,introduces the core architecture of Hadoop and the operating process and mechanism,and also introduces a smart medical platform system based on cloud computing.At the same time,it introduces the Apriori algorithm in the traditional algorithm in detail,pointing out its shortcomings,An improved Apriori medical data mining algorithm for medical cloud data based on Hadoop is proposed.In the mining algorithm based on association rules,Apriori technology,mining frequent itemsets and interesting associations in the transaction database,is not only the earliest used association rule mining technology,but also the most popular technology.After research,it is found that the traditional Apriori algorithm has two main bottlenecks: frequent scanning of the database;and the generation of a large number of candidate sets.In response to the inherent defects of the Apriori algorithm,some related improvements have been made: first,a new database mapping method is used to avoid repeated scanning of the database;then the frequent item sets and candidate sets are further pruned to improve the joining efficiency;the overlapping strategy is used to calculate support degree to achieve high efficiency.Finally,the improved algorithm will be simulated on the Hadoop platform,and the improved algorithm will be related to the traditional Apriori algorithm and the BITXOR algorithm introduced for comparison.The experimental results will be fully compared and analyzed.The experiment proves the improvement proposed in this paper Compared with the traditional and other improved algorithms,the algorithm has higher efficiency and superiority.