当前位置: X-MOL 学术J. R. Stat. Soc. A › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Willem van Zwet, 1934–2020
The Journal of the Royal Statistical Society, Series A (Statistics in Society) ( IF 2 ) Pub Date : 2020-10-05 , DOI: 10.1111/rssa.12602
P. J. Bickel , N. I. Fisher

Willem van Zwet passed away on July 2nd, 2020, after a short illness. He was an outstanding researcher, teacher and educator in mathematical statistics and probability, the father of statistics in the Netherlands, and the leading figure in opening up dialogue with colleagues in eastern Europe. His lasting contributions to the profession through his research and his work with professional societies, scientific journals and creating and developing institutions were recognized in numerous honours, including the AkzoNobel prize, the van Dantzig Award, the Wald Lectures of the Institute of Mathematical Statistics, a Humboldt Prize, an Honorary Doctorate from Charles University, memberships in the Royal Netherlands Academy of Sciences and the Academia Europe, and Knighthood in the Order of the Dutch Lion. An elegant and witty scholar equally at home in Dutch and English, he will be missed by a large community of friends and colleagues.

Willem Rutger van Zwet was born on March 31st, 1934, in Leiden. His father Jan van Zwet was a leading lawyer and his mother Leonie Sara Schot taught classical languages. After a somewhat erratic and varied schooling resulting from an intervening war and a certain self‐admitted idleness, his mathematics teacher advised him that he was not good enough to study mathematics but could possibly become a reasonably competent engineer. A similar attitude as an undergraduate at the University of Leiden saw him take 5 1 2 years to complete a 3‐year degree programme that started in physics and ended in mathematics. From there another erratic trajectory found him in the statistics group of the Mathematics Centre at Amsterdam. At the same time he attended lectures by David van Dantzig, and began a doctoral degree in mathematics at the University of Amsterdam with van Dantzig's successor, Jan Hemelrijk, eventually graduating with a highly original and largely self‐supervised doctoral thesis focused on non‐parametric statistics. After a short stint in the USA, he took up an appointment as Associate Professor of Statistics in the Mathematics Department at the University of Leiden in 1965, being promoted to Professor in 1968. Shortly thereafter he began a series of research visits to the University of California at Berkeley where he came under the influence of Erich Lehmann and began collaborating with faculty members there. Apart from frequent sabbatical and other visits to the USA, principally to the University of California at Berkeley and the University of North Carolina at Chapel Hill, he remained at Leiden for the rest of his academic career.

Bill, as he was known by most outside the Netherlands, made several very influential scientific contributions. Almost all were collaborative, but he always played a critical leading role. From the beginning, his focus was on non‐parametric models and methods. He realized early on that symmetric functions of independently and identically distributed variables were the basic elements of non‐parametric statistics including, as they did, rank and other permutation test statistics, Bradley Efron's bootstrap and Wassily Hoeffding's U‐statistics. Although the zeroth (law of large numbers), first (central limit theorem) and second (Edgeworth expansions) asymptotic theorems for means were classical in probability theory, knowledge of the behaviour of non‐linear symmetric statistics and their inferential consequences was more fragmentary. Hoeffding, through his expansion (later called analysis‐of‐variance expansion by Efron and Charles Stein) and, in particular, U‐statistics, and his and E. J. G. Pitman's parallel work on permutation tests had created some basic foundations.

Spurred on by novel results of Joe Hodges and Erich Lehmann, Bill initiated and led a collaborative study involving Peter Bickel, Friedrich Goetze and others, of the second‐order behaviour of one‐ and two‐sample rank tests. This was followed by ever more general work on Berry–Esseen bounds and Edgeworth expansions for U‐statistics. The final definitive results for symmetric statistics were in his Wald Lectures (1992) and in a beautiful paper on the Berry–Esseen bound for general symmetric statistics (1984). This work could be viewed in part as the most significant modern application of the Hoeffding expansion, which is most appropriate since Hoeffding was Bill's main scientific hero. In parallel, also with Goetze, Bickel and other collaborators, he made substantial, connected contributions to our understanding of resampling and the bootstrap.

More generally, Bill was intrigued by and continued to revisit ‘hard’ problems that were not necessarily related to the central theme of his research: a conjecture of Kakutani, his thesis on orderings of distributions initially inspired by probability plots; and his work on the contact process. A major feature that was apparent in all his work with students and collaborators was his love of working with other people at a high level. He was an excellent teacher and generous with his ideas, and his writing—for the most part in English—was distinguished by its wit, lucidity and elegance. Former students and ‘grand‐students’ hold leading academic positions throughout Holland and elsewhere. As a scholar, he received numerous invitations to visit, not least from the Mathematical Research Institute in Oberwolfach, where he may hold the record for visits (about 40).

In his wider professional life, Bill contributed on almost every front. He was one of the motive forces behind the creation of the Bernoulli Society. He served as President of the Bernoulli Society, the Institute of Mathematical Statistics and the International Statistical Institute. He was Editor‐in‐Chief of the journal Bernoulli, and the last person to be sole Editor‐in‐Chief of the Annals of Statistics, in which activity he received great assistance from his wife, Lucie. Perhaps most notable among his professional contributions were his initiatives to connect scholars in eastern Europe with colleagues in the west: working with Jim Durbin and Joe Gani, he helped to establish effective scientific relations between the European Region of the Institute of Mathematical Statistics and the probabilists and statisticians of eastern Europe, in particular by holding an Institute of Mathematical Statistics meeting in Budapest in 1972. He also revived the European Meeting of Statisticians to support interactions of European statisticians and gave many scientific visitors and refugees temporary refuge in the Netherlands through grants. This included personal ‘rescues’ of young individuals whom he led to very successful scientific careers, in particular, the students of Jaroslav Hájek at the Charles University in Prague, who had been left isolated by Hájek's untimely death.

Another of Bill's enduring initiatives has been Eurandom, an analogue to Oberwolfach, the US Mathematics Institutes such as the Mathematical Sciences Research Institute and the Isaac Newton Institute. He secured 5‐year start‐up funding and acted as founding Scientific Director during 1997–2000. Its initial focus was to recruit and train talented young researchers and to help them to find their way to tenured positions in academia and industry. Currently, it pursues a mission of running an extensive workshop and visitor programme for researchers.

Outside his professional life, Bill had diverse interests, ranging from ancient Australian Aboriginal culture, to Dixieland jazz (a passion shared with his two sons, Jan Rutger, a historian, and Erik, a statistician) and to the visual arts. A visit to him usually included the Mauritshuis and the Kroeller–Mueller Museum, as well as, often, the Rijksmuseum and van Gogh museums in Amsterdam. An enthusiastic table tennis player early in life, he much enjoyed hiking in the countryside later on. And he was always a splendid and entertaining companion when it came to the social aspects of conferences and workshops.

He is survived by his wife of 60 years, Lucie, his two sons and two grandchildren; and the memories of numerous people who had the good fortune to have their lives intersect with his.



中文翻译:

威廉·范·泽特(Willem van Zwet),1934–2020

Willem van Zwet在短暂患病后于2020年7月2日去世。他是数学统计学和概率论的杰出研究者,老师和教育家,荷兰的统计之父,并且是与东欧同事开展对话的主要人物。他通过研究以及与专业协会,科学期刊以及创建和发展机构的合作对专业做出的持久贡献获得了无数荣誉,包括阿克苏诺贝尔奖,范丹兹格奖,数学统计研究所的Wald演讲,洪堡奖,查尔斯大学的名誉博士学位,荷兰皇家科学院和欧洲科学院的会员资格以及荷兰狮勋章骑士勋章。

威廉·鲁特格·范·泽特(Willem Rutger van Zwet)于1934年3月31日出生在莱顿。他的父亲Jan van Zwet是首席律师,其母亲Leonie Sara Schot教授古典语言。由于一场战争和某种程度的自发性的闲暇而导致的学业有些不稳定和变化后,他的数学老师告诫他,他不足以学习数学,但有可能成为一名称职的工程师。与莱顿大学的本科生类似的态度使他接受了 5 1个 2 年完成3年制学位课程,该课程从物理学开始,以数学结束。从那里,另一条不稳定的轨迹在阿姆斯特丹数学中心的统计小组中找到了他。同时,他参加了David van Dantzig的演讲,并与van Dantzig的继任人Jan Hemelrijk一起在阿姆斯特丹大学获得数学博士学位,最终以高度原创且主要由自我监督的博士学位论文毕业,重点是非参数统计。在美国短暂工作后,他于1965年被任命为莱顿大学数学系统计学副教授,并于1968年晋升为教授。此后不久,他开始对加州大学伯克利分校进行一系列研究访问,在那里他受到埃里希·莱曼(Erich Lehmann)的影响,并开始与那里的教职员工进行合作。除了经常对美国进行休假和其他访问之外,主要是访问加州大学伯克利分校和北卡罗来纳大学教堂山分校,他在莱顿度过了余下的学术生涯。

比尔(Bill)是荷兰以外的大多数人所熟知的,他做出了几项非常有影响力的科学贡献。几乎所有人都是协作的,但他始终扮演着关键的领导角色。从一开始,他就专注于非参数模型和方法。他很早就意识到,独立且分布均匀的变量的对称函数是非参数统计的基本要素,包括秩和其他置换检验统计,Bradley Efron的bootstrap和Wassily Hoeffding的U统计。尽管均值的第零(大数定律),第一(中心极限定理)和第二(渐近扩展)渐近定理在概率论中是经典的,但非线性对称统计量的行为及其推论结果的知识更加零碎。霍夫丁,

在乔·霍奇斯(Joe Hodges)和埃里希·莱曼(Erich Lehmann)的新颖结果的刺激下,比尔发起并领导了一项涉及彼得·比克尔,弗里德里希·格茨(Friedrich Goetze)等人的协作研究,涉及一阶和二样本秩次检验的二次行为。接下来是关于Berry-Esseen边界和Edgeworth扩展U统计的更一般的工作。对称统计的最终确定结果在他的Wald Lectures中(1992年)和一篇关于Berry-Esseen关于一般对称统计的界的精美论文(1984年)。这项工作可以部分视为霍夫丁扩展的最重要的现代应用,这是最合适的,因为霍夫丁是比尔的主要科学英雄。同时,他还与Goetze,Bickel和其他合作者一起,为我们对重采样和自举的理解做出了重要的联系。

更笼统地说,比尔对并不一定与他的研究中心主题有关的“困难”问题很感兴趣,并继续对其进行研究:一个关于角谷的猜想,他的论题最初是由概率图启发而来的;和他在联系过程中的工作。在与学生和合作者的所有工作中,一个明显的主要特点是他热爱与高层合作。他是一位出色的老师,并且思想宽宏,而且他的写作(大部分是英语)以其机智,明晰和优雅而著称。以前的学生和“大学生”在荷兰和其他地方均担任主要学术职务。作为学者,他收到了许多邀请,尤其是来自上沃尔法赫(Oberwolfach)数学研究所的访问,

在他更广泛的职业生涯中,比尔几乎在每个方面都做出了贡献。他是伯努利学会创立的动力之一。他曾担任伯努利学会,数理统计研究所和国际统计研究所的主席。他是《伯努利》杂志的主编,并且是《统计年鉴》唯一的唯一主编在这项活动中,他得到了妻子露西(Lucie)的大力帮助。在他的专业贡献中,最著名的也许就是他倡议将东欧的学者与西方的同事联系起来:与吉姆·德宾和乔·加尼一起工作,他帮助建立了欧洲数理统计研究所与概率论者之间的有效科学关系。东欧和统计学家,尤其是1972年在布达佩斯举行的一次数学统计研究所会议。他还复兴了欧洲统计学家会议,以支持欧洲统计学家的互动,并通过赠款给予了荷兰许多科学访问者和难民临时避难所。其中包括年轻人的个人“救援”,他领导了非常成功的科学事业,尤其是,

比尔的另一项持久举措是与上奥伯法赫类似的Eurandom,美国数学研究所(如数学科学研究所和艾萨克·牛顿研究所)。他获得了5年的启动资金,并在1997-2000年期间担任科学总监一职。它最初的重点是招募和培训有才华的年轻研究人员,并帮助他们找到进入学术界和工业界的终身职位的途径。目前,它的使命是为研究人员举办广泛的研讨会和访问者计划。

在他的职业生涯之外,Bill的兴趣广泛,从古老的澳大利亚原住民文化到Dixieland爵士乐(与他的两个儿子,历史学家Jan Rutger和统计学家Erik共同拥有的激情)以及视觉艺术。对他的访问通常包括Mauritshuis和Kroeller-Mueller博物馆,以及阿姆斯特丹的国立博物馆和梵高博物馆。自幼热衷于乒乓球运动,后来他非常喜欢在乡下远足。在会议和研讨会的社交方面,他始终是出色而有趣的同伴。

他有60岁的妻子露西(Lucie),他的两个儿子和两个孙子,幸存下来。以及许多有幸与他相处的人的回忆。

更新日期:2020-10-06
down
wechat
bug