摘要
维修费用是舰船装备费用的重要组成部分,维修费用和影响费用的驱动因子之间直接的关系式很难确定,运用支持向量机回归可以很好的做到对维修费用的建模预测。为了解决支持向量机计算速度慢,效率低的问题,提出灰色关联度理论对费用相关费用驱动因子进行预处理,测试结果表明提高了SVM的效率,对SVM的泛化能力也有所提高。
The maintain expenses constitute the main part of the expenses of the ship＇s equipment.It＇s difficult to make sure the formula between the expenses and the factors which influence the expenses.It＇s better for us to use support vector machine（SVM） to estimate the expenses.But the efficiency of the SVM is low.In order to for solve the problem,the gray connection to process the factors data before estimating is applied.The test result indicates that the method can improve the efficiency of SVM,and also can improve the ability of SVM to estimate.
出处
《计算机与数字工程》
2010年第10期15-18,共4页
Computer & Digital Engineering
关键词
支持向量机
灰色理论
维修
费用预测
support vector machine
gray connection
maintain
expenses estimated
