在因果图理论中,采用了图形化和直接因果强度来表达知识和因果关系,它克服了贝叶斯网的一些不足,已发展成了一个能够处理离散变量和连续变量的混合模型.但是因果图的结构得由领域专家给出,这在实际中很难办到.鉴于因果图结构的复杂度随...在因果图理论中,采用了图形化和直接因果强度来表达知识和因果关系,它克服了贝叶斯网的一些不足,已发展成了一个能够处理离散变量和连续变量的混合模型.但是因果图的结构得由领域专家给出,这在实际中很难办到.鉴于因果图结构的复杂度随论域中节点个数的增加呈指数上升,寻找最有可能因果图网络结构成为了NP-HARD难题.文中给出了如何利用已知数据集,寻找最有可能的因果图网络结构设计的遗传算法(Genetic A lgorithm,GA).展开更多
An integrated, efficient and effective performance management system, "dynamic performance management system", is presented, which covers the entire performance management process including measures design, analysis...An integrated, efficient and effective performance management system, "dynamic performance management system", is presented, which covers the entire performance management process including measures design, analysis, and dynamic update. The analysis of performance measures using causal loop diagrams, qualitative inference and analytic network process is mainly discussed. A real world case study is carried out throughout the paper to explain how the framework works. A software tool for DPMS, Performance Analyzer, is also introduced.展开更多
文摘在因果图理论中,采用了图形化和直接因果强度来表达知识和因果关系,它克服了贝叶斯网的一些不足,已发展成了一个能够处理离散变量和连续变量的混合模型.但是因果图的结构得由领域专家给出,这在实际中很难办到.鉴于因果图结构的复杂度随论域中节点个数的增加呈指数上升,寻找最有可能因果图网络结构成为了NP-HARD难题.文中给出了如何利用已知数据集,寻找最有可能的因果图网络结构设计的遗传算法(Genetic A lgorithm,GA).
文摘An integrated, efficient and effective performance management system, "dynamic performance management system", is presented, which covers the entire performance management process including measures design, analysis, and dynamic update. The analysis of performance measures using causal loop diagrams, qualitative inference and analytic network process is mainly discussed. A real world case study is carried out throughout the paper to explain how the framework works. A software tool for DPMS, Performance Analyzer, is also introduced.