Tracy-Widom分布及其应用

Tracy-Widom Law with Applications

  • 摘要: Tracy和Widom 1990年代在研究高维随机矩阵特征根时发现一种新型分布, 文献中普遍称为Tracy-Widom分布, 它用于描述极值特征根的渐近性质. 随后二十多年的研究表明, Tracy-Widom分布如同经典的正态分布那样具有普适性, 适用于各种极值型随机现象. 作为例子, 本文简要描述了九种常见的随机模型, 它们都以某种方式与Tracy-Widom分布有关. 与正态分布相比较, Tracy- Widom分布的密度函数、分布函数、数字特征都显得非常复杂, 为了进一步推广和应用, 需要相当好的数学基础和计算能力. 但是, 由于该分布的重要性, 无论如何值得更多的关注.

     

    Abstract: Tracy and Widom found a new type of probability distribution in the study of high dimensional random matrices in the 1990s, which is nowadays normally called Tracy-Widom distribution. It is used to described the limiting distribution of the extremal eigenvalues in Gaussian Unitary Ensemble. Later on, the study in the past two decades indicates that Tracy-Widom distribution is universal like normal distribution and can be well used to describe a lot of seemingly distinct random phenomena. As illustrations, the paper briefly review nine widely studied random models, each of which is more or less related to Tracy-Widom distribution. Compared to normal distribution, Tracy-Widom distribution has horribly intricate distribution function, density function and moments. people need to use deep mathematical knowledge and advanced computation technology in order to extend and to apply Tracy-Widom distribution in practice. But it is absolutely worthy further study on account of its importance.

     

/

返回文章
返回