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
The recent financial crisis highlights the inherent weaknesses of the financial market. To explore the mechanism that maintains the financial market as a system, we study the interactions of U.S. financial market from...The recent financial crisis highlights the inherent weaknesses of the financial market. To explore the mechanism that maintains the financial market as a system, we study the interactions of U.S. financial market from the network perspective. Applied with conditional Granger causality network analysis, network density, in-degree and out-degree rankings are important indicators to analyze the conditional causal relationships among financial agents, and further to assess the stability of U.S. financial systems. It is found that the topological structure of G-causality network in U.S. financial market changed in different stages over the last decade, especially during the recent global financial crisis. Network density of the G-causality model is much higher during the period of 2007-2009 crisis stage, and it reaches the peak value in 2008, the most turbulent time in the crisis. Ranked by in-degrees and out-degrees, insurance companies are listed in the top of 68 financial institutions during the crisis. They act as the hubs which are more easily influenced by other financial institutions and simultaneously influence others during the global financial disturbance.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).
基金Supported by the National Natural Science Foundation of China under Grant Nos.7110317971102129+1 种基金11121403by Program for Young Innovative Research Team in China University of Political Science and Law
文摘The recent financial crisis highlights the inherent weaknesses of the financial market. To explore the mechanism that maintains the financial market as a system, we study the interactions of U.S. financial market from the network perspective. Applied with conditional Granger causality network analysis, network density, in-degree and out-degree rankings are important indicators to analyze the conditional causal relationships among financial agents, and further to assess the stability of U.S. financial systems. It is found that the topological structure of G-causality network in U.S. financial market changed in different stages over the last decade, especially during the recent global financial crisis. Network density of the G-causality model is much higher during the period of 2007-2009 crisis stage, and it reaches the peak value in 2008, the most turbulent time in the crisis. Ranked by in-degrees and out-degrees, insurance companies are listed in the top of 68 financial institutions during the crisis. They act as the hubs which are more easily influenced by other financial institutions and simultaneously influence others during the global financial disturbance.
