Research On CVD Prediction Based On SVM Optimized By Hybrid Meta-heuristic Algrithm

Cardiovascular disease(CVD)is currently the chronic disease with the highest fatality rate in the world.Therefore,the study of Cardiovascular Disease is also an urgent topic.Many scholars have devoted themselves to researching a high-quality prediction method as a tool to assist medical examination and diagnosis.Through the reading of a large number of literatures and reference,I find that the construction of prediction models for Cardiovascular Diseases mainly focuses on the categories of machine learning such as Decision Tree algorithm and SVM model.The purpose of this study is to improve the accuracy and reliability of cardiovascular disease prediction,and what algorithm model is used to achieve this goal.I take Coronary Heart Disease among Cardiovascular Diseases as the research object,collecting patient diagnostic data in local hospitals,and carrying out the full-text research based on the SVM model in machine learning and the research method of improving model parameters.This paper starts with the data of a local authoritative tertiary hospital,through data processing,feature selection,model construction and parameter optimization,to make an in-depth analysis of the prediction of Cardiovascular Disease.Firstly,the related theory is introduced to SVM,GA algorithm and PSO algorithm.Then,the grid search method,GA algorithm,PSO algorithm and the GA-PSO hybrid meta-heuristic algorithm proposed in this paper are used to establish the SVM parameter optimization prediction model,and several Benchmark test functions are used to compare the accuracy and convergence effect of several algorithms.The results also verify that the performance of the GA-PSO hybrid algorithm is better than the GA algorithm,and the GA algorithm is better than the PSO algorithm.Then the above model is applied to the data of the research object to compare which has better reliability.The analysis results show that the improved GA-PSO algorithm has the highest prediction accuracy,followed by GA,PSO and grid search.The prediction accuracy of cardiovascular disease obtained by GA-PSO hybrid meta-heuristic algorithm on SVM parameter optimization is 87.75%,while the model prediction accuracy using GA algorithm,PSO algorithm and grid search optimization is 84.27%,83.81% and 79.6% respectively.Finally,P value,R value,F1 value and AUC value of ROC curve are used to evaluate the model,which strengthened the reliability of the improved GA-PSO hybrid meta-heuristic algorithm.The improved GA-PSO hybrid meta-heuristic algorithm proposed in this paper has better prediction results in Cardiovascular Diseases,and has reference value in assisting doctors to make more accurate judgments of Cardiovascular Diseases.