基于全局信息的动态激励评价方法及激励策略  

Dynamic incentive evaluation based on overall information and incentive strategy

作　　者：郭亚军[1] 周莹[1] 易平涛[1] 李伟伟[1] 

机构地区：[1]东北大学工商管理学院,辽宁沈阳110819

出　　处：《系统工程学报》2017年第2期282-288,共7页Journal of Systems Engineering

基　　金：国家自然科学基金资助项目(71071030;71071031);国家自然科学基金面上资助项目(71671031)

摘　　要：针对动态激励评价问题,提出了一种诱导被评价对象个体行为向评价者全局期望动态趋同的评价方法.首先,通过对全局被评价对象的观测值与变化量进行"激励型分层"得到具有奖优惩拙思想的"全局激励系数",据此在动态环境中集结出带有激励作用的被评价对象综合评价值与排序.然后,运用一个算例对该方法的可行性与有效性进行了说明.此外,通过对激励策略进行仿真,分析不同策略下相对初始排序的变化程度及不同策略间的变动对排序结果的影响,从而为评价者制定激励型分层策略提供理论依据.For the dynamic incentive evaluation problem, this paper proposes an evaluation method that in- duces the evaluation object＇s behavior convergence to the evaluator＇s expectation. First, an overall incentive coefficient is derived which rewards excellent indexes and punishes worse indexes by classifying the incentive layer of the overall observed value and variation of the overall evaluation object. Then an example is used to illustrate the feasibility and validity of this method. In addition, through a simulation of the incentive strategy, the effects of different strategies on the initial sorting and effects of the change of each strategy on sorting are analyzed. Evaluators can develop incentive layered strategies according to this studies.

关 键 词：综合评价 动态评价 信息集结 激励型分层 随机模拟 

分 类 号：C934[经济管理—管理学]