Poisson distribution

When the probability of A is p and the probability of not A is 1‐p in a certain population, if n pieces are randomly taken from the popuplation, the probability of A being x is _nC_xp^x(1-p)^n-x, which is called the binominal distribution. When the probability p is very small, the binominal distribution becomes "Poisson distribution" e^-λ・λ^x/x !. In the case of TEM, the Poisson distribution is applied to the probability of inelastic scattering events. The higher-order plasmon scattering intensities in an EELS spectrum are removed by assuming Poisson distribution. Poission distribution is also effectively used for evaluation of counting errors in a CCD detector.