The Research On The Prediction Of Cardiovascular Diseases Based On Random Forest And Support Vector Machine

Nowadays,cardiovascular disease is a prevalent chronic disease that has a serious impact on the health of patients and brings the heavy medical burden to the society and patients.A quick and effective prediction of cardiovascular disease helps patients to discover the disease timely.The time cost of the patients' disease information is various in the context of rapid prediction and classification of patients' condition,so the division of the patients' disease information according to the time cost and the effective use of the disease information are of great significance to the prediction of cardiovascular disease.The current main research is mainly focused on two aspects.One is about the establishment of traditional cardiovascular disease risk assessment models,this kind of research requires a large number of patients as long-term follow-up subjects,is not suitable for the rapid and efficient research.The other is the application of machine learning method in the feature selection and prediction of cardiovascular disease with abstracting the cardiovascular disease prediction into a machine learning problem.However,the current research has not considered more about the context of quick and efficient prediction of cardiovascular disease.Therefore,in view of the characteristics of the prediction of cardiovascular disease,this paper used the machine learning method to construct the cardiovascular disease prediction model with the consideration of the time cost,and predict the cardiovascular disease quickly and efficiently.Firstly,the paper gave the combination of the cardiovascular disease risk factors for the different stages based on the time cost.Secondly,in order to utilize the patients' disease information efficiently,the random forest method was applied to eliminate the irrelevant and redundant information from the cardiovascular disease data,and identified the key cardiovascular disease risk factor in the process of feature selection.Finally,the SVM(Support Vector Machine)was used to do the prediction of cardiovascular disease with the identified key risk factor of each stage,the genetic algorithm was used to optimize the internal parameter of SVM to improve the prediction accuracy,and the prediction results of cardiovascular disease under the combination of risk factors at each stage with the different time cost was obtained and evaluated,the evaluation of the prediction results provided the reference and suggestion for the quick classification and prediction of cardiovascular disease.At last,the paper used the above proposed model to predict the coronary heart disease.The prediction results of the proposed model were compared with the prediction results of KNN(K-nearest neighbor)algorithm,Logistic regression and BP neural network,the better performance verified the efficiency of the proposed model.