王骏, 陈萍. 一种基于多维尾条件期望的扩散模型检验方法[J]. 应用概率统计, 2018, 34(2): 135-144. DOI: 10.3969/j.issn.1001-4268.2018.02.002
引用本文: 王骏, 陈萍. 一种基于多维尾条件期望的扩散模型检验方法[J]. 应用概率统计, 2018, 34(2): 135-144. DOI: 10.3969/j.issn.1001-4268.2018.02.002
WANG Jun, CHEN Ping. A Test for Diffusion Model Based on Multi-Dimensional Tail Condition Expectations[J]. Chinese Journal of Applied Probability and Statistics, 2018, 34(2): 135-144. DOI: 10.3969/j.issn.1001-4268.2018.02.002
Citation: WANG Jun, CHEN Ping. A Test for Diffusion Model Based on Multi-Dimensional Tail Condition Expectations[J]. Chinese Journal of Applied Probability and Statistics, 2018, 34(2): 135-144. DOI: 10.3969/j.issn.1001-4268.2018.02.002

一种基于多维尾条件期望的扩散模型检验方法

A Test for Diffusion Model Based on Multi-Dimensional Tail Condition Expectations

  • 摘要: 为了解决多维扩散模型的设定检验问题,我们基于多维尾条件期望~(CTE)~制定了一个检验统计量.虽然不能直接估计出多维扩散模型的转移密度矩阵, 但是可以估计出每个分量的转移密度, 再通过~CTE~将每个分量联合起来,建立了一个真正的多维统计量. 最后通过仿真模拟评估了该检验方法的检验的性能.

     

    Abstract: In order to solve the problem of testing multidimensional diffusion models, we develop a test statistic based on Multi-Dimensional Tail Condition Expectations (CTEs). Although it is almost impossible to estimate the transition density matrix of a multidimensional diffusion model directly, the transition density of each component can be estimated and each component can be combined by the CTE to establish a true multidimensional statistics. Finally, the performance of the test is evaluated through simulation.

     

/

返回文章
返回