Application Of SVM Classifier Based On Parameter Optimization In The Diagnosis Of Secondary Sjogren's Syndrome

Siccasyndrome is an exocrine gland of autoimmune chronic diseases,usually divided into primary Sjogren's syndrome and secondary Sjogren's syndrome,secondary Sjogren's syndrome often secondary to Systemic Lupus Erythematosus(abbreviation:SLE)or other diseases,so it is easy to be ignored.In order to solve the problem that the treatment of.SLE patients complicated with secondary Sjogren's syndrome is not easy to be diagnosed and treated in the course of treatment,a new idea of combining SVM with early diagnosis of systemic lupus erythematosus secondary Sjogren's syndrome was proposed,the main idea is to intelligent machine algorithm SVM classifier applied to the actual data classification.In this dissertation,141 cases of patients with the case of the study,through the data screening process,respectively,using the cross-verification method,grid search method,the standard PSO algorithm respectively optimize for SVM model classifier,the penalty parameter C and kernel function parameter g,and finally found that the PSO algorithm SVM parameter model not only classify the best results,its generalization ability also greatly improved.Then,a new method of PSO is proposed to improve the particle swarm optimization algorithm.Finally,the four kinds of optimization methods are programmed by MATLAB software,and the results are expressed in the workspace window in an intuitive image to express the final classification accuracy of SVM model classifier.The accuracy of the final classifications was 82.3529%,88.2353%,90.1961%and 92.1569%,respectively,for SLE patients with secondary Sjogren's syndrome.The results of the final results show that the optimization of support vector machine classification based on the improved particle swarm optimization algorithm based on chaos mechanism is more scientific and rigorous than the cross validation and grid search method.Particle swarm optimization(SVM)model,it can be clearly seen that the improved particle swarm optimization algorithm based on chaotic mechanism can improve the accuracy of SVM classifier to classify the diagnosis of SLE secondary Sjogren's syndrome.