基于GWO-SVR的冠心病住院费用预测

Prediction of hospitalization costs for coronary heart disease based on GWO-SVR

冠心病,作为世界上威胁中老年人健康最常见的疾病之一,近年来诊疗费用不断攀升。因此对冠心病住院费用进行准确的预测,对于着力控制其医疗费用增长具有重要意义。本文运用灰狼优化算法(Grey Wolf Optimizer,GWO)对支持向量机回归(Support Vector Regression,SVR)模型的惩罚系数C和核函数方差g进行优化,实现了基于GWO-SVR的冠心病住院费用预测模型。研究结果表明,相较于原始SVR模型,差分进化算法(Differential Evolution,DE)、布谷鸟搜索算法(Cuckoo Search,CS)、粒子群算法(Particle sw arm optimization,PSO)优化的SVR模型,灰狼优化算法可以在最短时间内实现参数优化,并且能更加精准有效的预测出冠心病住院费用变化的趋势。

Coronary heart disease(CHD),as one of the most common diseases threatening the health of the middle-aged and elderly people in the world,has been increasing the cost of diagnosis and treatment in recent years.Therefore,accurate prediction of the hospitalization cost of coronary heart disease is of great significance for controlling the growth of medical expenses.In this paper,the grey Wolf optimization algorithm is used to optimize the punishment coefficient C and kernel variance G of the SVM regression model,so as to realize a gwO-SVR based hospital cost prediction model for coronary heart disease.The results show that,compared with the SVR parameter models optimized by differential evolution algorithm(DE),Cuckoo search algorithm(CS) and particle swarm optimization(PSO),the grey Wolf optimization algorithm can realize parameter optimization in the shortest time,and can more accurately and effectively predict the trend of changes in the hospitalization costs of coronary heart disease.