摘要
针对商业银行中各利益相关者对银行资本配置的影响较为复杂,我国商业银行相关数据缺乏等问题,文章以支持向量机回归模型建立我国股份制商业银行各利益相关者对银行资本配置影响的仿真模型,并利用我国股份制商业银行1999～2006年的数据进行了仿真计算,并同PLS回归、BP神经网络的预测结果进行了比较,证明这一思路是可行、有效的。
Because the influences of stakeholders on capital allocation are more complicated,and the data related to commercial banks are scarce,the paper established the simulation model of forecasting the influences of stakeholders on capital allocation in joint-stock commercial banks using support vector machine regression,and utilized the samples of Chinese joint-stock commercial banks during the period of 1999 ～ 2006 to simulate.Comparing the result of SVM with that of PLS regression and BP neural network,the paper demonstrated that the idea was feasible and effective.
出处
《管理工程学报》
CSSCI
北大核心
2010年第4期148-154,共7页
Journal of Industrial Engineering and Engineering Management
基金
国家自然科学基金资助项目(70972053)
陕西省重点学科建设专项资金资助项目(107-00X902)
陕西省社会科学基金项目(09E026)
