John C. Gittins
John C. Gittins
|Born||1938 (aged 80)|
|Alma mater||University of Cambridge (MA), University of Wales (PhD)|
|Known for||Gittins index|
|Awards||Rollo Davidson Prize (1982)|
Guy Medal (Silver, 1984)
|Thesis||Optimal resource allocation in chemical research (1968)|
|Doctoral advisor||Dennis Lindley|
He is renowned as the developer of the "Gittins index", which is used for sequential decision-making, especially in research and development in the pharmaceutical industry. He has research interests in applied probability, decision analysis and optimal decisions, including optimal stopping and stochastic optimization.
Gittins was an Assistant Director of Research at the Department of Engineering, Cambridge University from 1967 to 1974. Then he was a lecturer at Oxford University from 1975 to 2005 and head of the Department of Statistics there for 6 years. In 1992, Oxford University awarded him the degree Doctor of Science (D. Sci.). In 1996 he became a Professor of Statistics at Oxford University.
- (1989) Multi-Armed Bandit Allocation Indices, Wiley. ISBN 0-471-92059-2
- (1985) (with Bergman, S.W.) Statistical Methods for Pharmaceutical Research Planning, CRC Press. ISBN 0-8247-7146-X
- (2000) (with H. Pezeshk) "How Large Should a Clinical Trial Be?", The Statistician, 49 (2), 177–187 )
- (2001) (with G. Harper) "Bounds on the Performance of a Greedy Algorithm for Probabilities". Mathematics of Operations Research, 26, 313–323
- (2003) "Stochastic Models for the Planning of Pharmaceutical Research", Journal of Statistical Theory and Applications, 2 (2), 198–214.
- (2011) (with K. D. Glazebrook and R. R. Weber) Multi-Armed Bandit Allocation Indices, second edition, Wiley, ISBN 0-470-67002-9
- Whittle, Peter (1980). "Multi-armed bandits and the Gittins index". Journal of the Royal Statistical Society, Series B. 42 (2): 143–149. JSTOR 2984953.
- Whittle, Peter (2002). "Applied probability in Great Britain (50th anniversary issue of Operations Research)". Oper. Res. 50 (1): 227–239. doi:10.1287/opre.126.96.36.19992. JSTOR 3088474.