Auxiliary Diagnosis Method of Chest Pain Based on Machine Learning

Abstract

Chest pain is sudden, its pathological causes are complex and various, fatal or non-fatal so that improving the diagnostic accuracy is extremely important in the emergency system of prehospital and hospitals. Therefore, we propose a method of introducing a decision tree, support vector machine, and KNN algorithm in machine learning into the auxiliary diagnosis of chest pain. First select the algorithm with better performance among decision tree, support vector machine, and KNN algorithm; Then compare the classification performance of the CART algorithm, the support vector machine using the Gaussian kernel function, and the K nearest neighbor algorithm using the Euclidean distance to select the best; Finally, through the analysis of the experimental results, the support vector machine algorithm with Gaussian kernel function is obtained. Its detection time and diagnosis accuracy rate are the best among the three algorithms, which can assist medical staff in the emergency system to carry out targeted chest pain diagnosis.