Piet Groeneboom

Petrus (Piet) Groeneboom (born 24 September 1941 ) is a Dutch statistician who made major advances in the field of shape-constrained statistical inference such as isotonic regression, and also worked in probability theory.

Education and career Groeneboom was born in Scheveningen.

Before and immediately after obtaining his master's degree, Groeneboom worked at the psychological laboratory of the University of Amsterdam. After his second stint there ended in 1973, he moved to the , which at the time was called (Mathematical Centre), in the same city. since his retirement in 2006. He has also held positions at the Vrije Universiteit Amsterdam and the University of Washington. Following his retirement he came to public attention for his statistical work in the retrial of Lucia de Berk, a Dutch nurse, who had been convicted of murder.

Research In 1979, Groeneboom, together with Oosterhoff and Frits H. Ruymgaart, formulated and proved Sanov's theorem in a finer topology than had been known at the time. A paper he published in 1983 on properties of Brownian motion gave rise to a large body of literature on minorants of more general stochastic processes. One of the main areas of work of Groeneboom has been shape constrained statistical inference, which includes isotonic regression, an area with links to the aforementioned minorant problems, In a 1985 article on an estimator of a monotone density named after Ulf Grenander, he introduced the switching (or switch) relation, which came to be used widely in the area. He found the analytic form of Chernoff's distribution, which later was understood to be omnipresent in monotone problems, in the 1980s,

Since the late 1980s, Groeneboom has also worked on censored regression models.

Together with Eric Cator, Groeneboom contributed to the probabilistic analysis of the Hammersley process, a continuous interacting particle system (IPS). Methods similar to theirs were subsequently applied to other IPSs. He is known to influence academic thought amongst some American statisticians such as Jon A. Wellner. They argued that statistical considerations that had led to the initial suspicions of murder, and those which had remained at the center stage of the case afterwards, were flawed. The effort was ultimately successful and de Berk was finally acquitted of all accusations in 2010.

Groeneboom is a fellow of the Institute of Mathematical Statistics, and an elected member of the International Statistical Institute. In 2013, he delivered the Wald lectures at the Joint Statistical Meetings in Montreal.

Books authored