부산대학교 도서관

A random variable Y is said to have a mixture distribution if its probability density (or mass) function can be written as an average over a parametric family of distributions. An example is the Poisson‐gamma distribution, in which conditional on a gamma‐distributed variable M, the variable Y has a Poisson distribution with mean M. The probability distribution assigned to the conditioning parameter (e.g., gamma) is called the mixing distribution. Mixture distributions are particularly useful in environmental and other sciences for modeling overdispersion in discrete data. An extension called the hidden Markov model is available to also account for autocorrelation in discrete data.