Arthur P. Dempster

from Wikipedia, the free encyclopedia
Arthur P. Dempster

Arthur Pentland Dempster (* 1929 ) is a professor emeritus at Harvard University and a well-known mathematical statistician .

Live and act

Dempster studied mathematics and physics at the University of Toronto and received his doctorate in 1956 with John Tukey on "The two-sample multivariate problem in the degenerate case" . Since 1958 he has been a professor at Harvard University. There he was Chair of the Department of Statistics from 1969 to 1975, 1977 to 1979 and 1982 to 1985.

Significant scientific contributions are made in the field of evidence theory , which is now often called Dempster-Shafer theory . The so-called Dempster rule of combination , which in some respects represents a generalization of Bayes' rule , plays a major role in this theory . Dempster is also significantly involved in the development of the well-known EM algorithm .


1962 Fellow of the American Statistical Association , 1963 Fellow of the Institute of Mathematical Statistics , 1967–68 Guggenheim Fellow, 1990–91 Fellow of the Center for Advanced Study in the Behavioral Sciences, 1997 Fellow of the American Academy of Arts and Sciences .

Web links

Individual evidence

  1. Dempster, AP (1967): Upper and lower probabilities induced by a multivalued mapping , Annals of Mathematical Statistics 38, 325–339 doi: 10.1214 / aoms / 1177698950
  2. Dempster, AP (2008): The Dempster-Shafer calculus for statisticians , International Journal of Approximate Reasoning 48, 365-377. doi: 10.1016 / j.ijar.2007.03.004
  3. ^ AP Dempster (1968): A generalization of Bayesian inference. Journal of the Royal Statistical Society. Series B 30, 205-247 full text
  4. Dempster, AP, Laird. NM, Rubin, DB (1977): Maximum-Likelihood from incomplete data via the EM algorithm (with discussion) . Journal of the Royal Statistical Society, Ser. B 39, 1-38 full text