A SIMPLE DERIVATION OF THE POSSION DISTRIBUTION
Abstract
One of the most important stochastic processes is the Poisson process, in which it is assumed that (a) the numbers of events occurring in nonoverlapping time intervals are independent; (b) the probability of one event's occurring during time dt is lambda dt + o(dt), where lambda is a constant, while the probability that two or more occur is o(dt). Using only the simplest kind of reasoning from probability theory, the Poisson distribution is deduced from the basic assumptions (a) and (b). Consequently, the need for viewing the Poisson distribution as a limiting case of some other distribution is obviated. In addition the technique used readily generalizes to the case in which lambda depends on t.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 11, 1953
- Accession Number
- AD0604222
Entities
People
- Robert E. Kalaba
Organizations
- RAND Corporation