Incremental Support Vector Machine Algorithm Integrated With Cloud Computing And Application Research

In recent years, with the development of health care reform, our country has a gooddevelopment opportunity in the large hospitals informatization and digitalization. Atpresent, a very primary reason of "the difficulty and costliness of medical service" is theseparate development of the medical industry information systems, with high operatingcosts, poor expansibility, difficulty maintenance and hard to achieve data sharing. Lots ofvaluable information has not been fully excavated and used. However, with the promotionof economy and culture level, people have increasingly high expectations for medicalservices. The research on how to improve the medical service level is very important toimprove the current situation of our country medical treatment.The Support vector machine is based on statistical theory. This method is superior toother existing methods and shows a lot of good performance, was regard as the optimaltheory in classification and regression problems of small samples set at present. Cloudcomputing is an emerging computing model for adaptation to increasing data, whoseadvantages are easy to develop and support large-scale data parallel processing.This paper will briefly discuss about China’s medical information status, incrementalsupport vector machine, cloud computation theory. Next analysis and improve existingincremental support vector machines algorithm, establish incremental support vectormachine algorithm integrated with cloud computing model based on these theories. Themodel makes use of advantage of cloud computing to solve the large-scale data, to solvethe problem of support vector machine can only be computed for small samples set. Andthis model is applied to the assisted diagnosis of Graves’ disease, and achieved goodresults.This paper will do the following research work:First, propose incremental support vector machines algorithm based on the internalrectangular, reduce incremental SVM training time, increase the prediction accuracy.Second, the support vector machine and cloud computing platform are combined toestablish incremental support vector machine algorithm integrated with cloud computingmodel. Research on this algorithm with the traditional support vector machine algorithmfor comparative analysis, study this model’s advantages. Choose one of the larger data fromUCI data sets, to test S-SVMCC algorithm and M-SVMCC algorithm performance in timeconsumption and efficiency. The S-SVMCC algorithm is not only quick calculation, and high prediction accuracy. Conclude that: the method has fast calculation speed and highprediction accuracy.Third, according to incremental growth characteristics of medical case data, choosethe suitable S-SVMCC algorithm. The algorithm is applied to the case of Graves diagnosis,and gets good result.