William Rae Thompson

William Rae Thompson was an American mathematician and statistician best known for inventing Thompson sampling, a heuristic for the multi-armed bandit problem that is now a standard technique in machine learning and online decision-making. He earned a bachelor's degree at Columbia University in 1923 and a doctorate in mathematics from Yale University in 1930, and he worked at Yale from 1924 to 1936, partly in the Department of Pathology. He later moved to the New York State Department of Health in Albany, where he published statistical analyses of medical testing and immunology. In 1933 he published the Biometrika paper that introduced the method, and in 1953 he became a Fellow of the American Statistical Association.