基于SVM的武器装备批量生产成本费用模型  

Cost model on batch manufacturing of weapon system using SVM

作　　者：薛峰[1] 高尚[1] 

机构地区：[1]江苏科技大学计算机科学与工程学院,江苏镇江212003

出　　处：《舰船科学技术》2013年第1期126-130,共5页Ship Science and Technology

基　　金：人工智能四川省重点实验室开放基金(2009RY001);"青蓝工程"(苏教师[2010]27号)

摘　　要：利用支持向量机建立了批量生产成本费用模型,批量生产成本费用与影响因素的关系可通过支持向量机的核函数和拉格朗日乘子得到体现。由于支持向量机综合考虑了结构风险和经验风险,使得模型的泛化能力增强。通过实例与熟练曲线法、最小二乘法和加权最小二乘法的结果进行了比较,结果表明支持向量机比较精确。A cost model on batch manufacturing of weapon system is established by using support vector machine. The relationship among cost on batch manufacturing with affect factors is described through kernel function and Lagrange coefficient of support vector machine. Because this method considers the struclural risk and empirical risk, it enhances the forecast precision of the model. The method is illustrated through examples. The comparing results show that the support vector machine method is more accurate than profidency curve method, least squares and weighted least squares.

关 键 词：系统评估与可行性分析 批量生产 支持向量机 加权最小二乘 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]