| Heart disease collectively represents various diseases about heart, including rheumatic heart disease, congenital heart disease, hypertensive heart disease, coronary heart disease, myocarditis, etc. As a disease of high mortality rate, heart disease has become the primary factor of human death and brought huge economic burden and life disaster to the sick family. In the medical field, if heart disease can be diagnosed and intervened at early stage, and patients can take effective treatment as early as possible, the sudden disastrous consequences of heart disease can be avoided. Therefore, carrying out physiological parameter monitoring on patients and effective studies on the diagnosis of heart disease is of great application value for the early intervention and treatment of heart disease.The thesis revolves around the heart disease diagnosis in the field of disease diagnosis; through the use of multiple physiological parameters monitoring of patients, advanced data analysis and artificial intelligence methods are combined; diagnosis model of multiple physiological parameters of heart disease is established based on SOA- SVM so as to effectively diagnose patients’ heart conditions and diagnose heart disease more practically and effectively. The main contents include the following aspects:Firstly, the selection and processing of multiple physiological monitoring parameters of heart disease. In the thesis, the mechanism and the status quo of cardiovascular disease are discussed and the study questions are identified; several physiological monitoring indicators of heart disease are analyzed and the data are preprocessed.Secondly, the study on the optimization method based on model parameter of SOA-SVM. In the application process, the selection of kernel function parameters and penalty factor of classification model of Support Vector Machine(SVM) is quite vital, since they affect the classification accuracy of the model. In the thesis, the heuristic random search algorithm of SOA is studied; the advantages of the PSO and GA algorithms are verified by the simulation analysis; a good foundation is laid for parameter optimization of SVM model and the establishment of optimal classification diagnosis model.Thirdly, the establishment and application of diagnosis model of heart disease based on SOA- SVM. Through the establishment of feature parameter monitoring of effective cases, the physiological parameters of heart patients can be described. The introduction of optimization algorithm SOA can efficiently optimize the model parameters of the SVM. In the thesis, the diagnosis model of multiple physiological parameters of heart disease based on the SOA- SVM is put forward, and the application process of the model is discussed. The effectiveness of the method is validated through the data sets of heart diseases provided by UCI. |
