The results show that the neural network model can fit these data well, with a total ppv of 0.798. On the other hand, the vector machine algorithm is supported to have a small mapping error and error to the fit of the data. The higher the hosmer-lemeshow fit quality test value, which indicates the advantages of supporting vector machine model in data, and provides better prediction for cad diagnostics. In addition, the support vector algorithm predicts the ppv and sensitivity of patients with coronary heart disease are higher than the artificial neural network model. Similarly, previous studies have shown that disease can be predicted using a support vector algorithm and distinguish patients from non-patients with greater accuracy
正在翻译中..