Comments on “A Gibbs Sampler for a Class of Random Convex Polytopes”

Comments on “A Gibbs Sampler for a Class of Random Convex Polytopes”
复制标题

对“一类随机凸多面体的吉布斯采样器”的评论

DOI:
10.1080/01621459.2021.1950002
复制
发表时间:
2021
影响因子:
3.7
通讯作者:
Zhang, Kai
Zhang, Kai
中科院分区:
数学1区
文献类型:
--
作者:
Hoffman, Kentaro;Hannig, Jan;Zhang, Kai

文献摘要

参考文献

相似文献

我们要祝贺作者对数理统计中这个长期悬而未决的问题做出的贡献。他们巧妙地实现了 MCMC,以获得多项分布参数的基于单纯形的 Dempster–Shafer (DS) 样本,及其与图论的联系非常发人深省。我们预计本文中看到的贡献将在未来几年产生影响。在我们的评论中,我们希望特别就各种 DS 方法在非参数独立性检验中的应用提出我们的想法。非参数依赖性检测是一个经典的统计问题,但由于其在机器学习中的应用,最近引起了统计学家和计算机科学家的极大兴趣。参见 Hoeffding (1948)、Székely、Rizzo 和 Bakirov (2007)、Gretton 等人 (2007)、Reshef 等人 (2011)、Heller、Heller 和 Gorfine (2013)、Chatterjee (2020)、Shi 等人 (2020) 以及其中的参考文献。非参数独立性检验的一种通用方法是多分辨率方法。请参阅 Ma 和 Mao (2019)、Zhang (2019)、Lee、Zhang 和 Kosorok (2019)、Gorsky 和 ​​Ma (2018)、Zhang、Zhao 和 Zhou (2021) 的一些近期作品。这种方法的一些优点包括统一一致性、最小最大最优功率、清晰的可解释性和高效的计算。
We would like to congratulate the authors for their contribution to this longstanding open problem in mathematical statistics. Their clever implementation of MCMC to obtain simplex-based Dempster–Shafer (DS) samples for parameters of multinomial distribution and its connection to graph theory is extremely thought provoking. We expect that the contributions seen in this article will have impacts for years to come. In our comment, we would like to contribute our thoughts particularly on the applications of the various DS approaches to nonparametric tests of independence. Nonparametric dependence detection is a classical statistical problem but recently gains great interest from both statisticians and computer scientists due to its applications in machine learning. See Hoeffding (1948), Székely, Rizzo, and Bakirov (2007), Gretton et al.(2007), Reshef et al.(2011), Heller, Heller, and Gorfine (2013), Chatterjee (2020), Shi et al.(2020), and references therein.One general approach in nonparametric tests of independence is the multiresolution approach. See some recent works by Ma and Mao (2019), Zhang (2019), Lee, Zhang, and Kosorok (2019), Gorsky and Ma (2018), and Zhang, Zhao, and Zhou (2021). Some advantages of this approach include uniform consistency, minimax optimal power, clear interpretability and efficient computation.
基于中心向外排序和符号的一致无分布独立性检验的速率最优性
DOI: --
发表时间: 2020
期刊:
影响因子: --
作者:
Hongjian Shi;M. Hallin;M. Drton;Fang Han
通讯作者: Fang Han
DOI: 10.1080/01621459.2020.1758115
发表时间: 2020-05-28
影响因子: 3.7
作者:
Chatterjee, Sourav
通讯作者: Chatterjee, Sourav