en.wikipedia.org/wiki/Rare_event_sampling
1 correction found
The first published rare event technique was by Herman Kahn and Theodore Edward Harris in 1951,
This date is too late. Rare-event Monte Carlo techniques were already published by Herman Kahn in 1950, and historical reviews trace the early published work to 1949–1950 rather than first appearing in 1951.
Full reasoning
The claim says the first published rare-event technique was by Kahn and Harris in 1951. But there is earlier published work.
- PubMed lists H. Kahn, "Random sampling (Monte Carlo) techniques in neutron attenuation problems--I." as a journal article published in May 1950.
- An IEEE Signal Processing Society historical overview of importance sampling states that the method was already used for "inference of rare events", specifically estimating the probability of nuclear particles that penetrate shields [Kahn, 1950].
- A historical review in the 2017 Winter Simulation Conference goes further and says early advances in efficient rare-event simulation were published in Kahn and Harris (1949), Kahn (1950, 1954), etc., and discusses Kahn (1950) as an early importance-sampling source.
So the 1951 Kahn–Harris publication was not the first published rare-event technique; published rare-event simulation work existed at least by 1950, and some historical reviews point to 1949 technical reports as well.
3 sources
- Random sampling (Monte Carlo) techniques in neutron attenuation problems--I. - PubMed
H KAHN. Publication date: 1950 May. Journal article: Random sampling (Monte Carlo) techniques in neutron attenuation problems--I.
- Adaptive Importance Sampling: The Past, the Present, and the Future | IEEE Signal Processing Society
"It was implicitly present in statistical physics for inference of rare events, and in particular for estimating the probability of nuclear particles that penetrate shields [Kahn, 1950]."
- History of Improving Statistical Efficiency (2017 Winter Simulation Conference)
The review states that early advances were 'published in Kahn and Harris (1949), Kahn (1950, 1954), and Kahn and Marshall (1953)' and discusses Kahn (1950) as an early importance-sampling source.