multinomial distribution


Also found in: Encyclopedia, Wikipedia.

mul·ti·nom·i·al dis·tri·bu·tion

probability distribution associated with the classification of each of a sample of individuals into one of several mutually exclusive and exhaustive categories.
Mentioned in ?
References in periodicals archive ?
k] |[eta]), and multinomial distribution samples topic word [x.
To estimate the expected frequencies and the model parameters, it is necessary to compare the estimated and expected values with the negative logarithmic likelihood based on multinomial distribution - ln L{L|[[mu].
In stochastic pooling, the pooled maps are determined by sampling from each pooling region using a multinomial distribution.
Additional material examines probability distributions, elements of set theory, multinomial distribution of single-voxel imaging, and other subjects.
The expression in Equation (1) is known as the multinomial distribution (Lawal, 2003).
The multinomial distribution (with fixed number of trials and unknown probabilities) in itself is an exponential family whose inverse parameter mapping is the generalization of the logistic function, called softmax function, that corresponds to equation (21).
But as N gets large, computing the probabilities associated with every element in the sample space of a multinomial distribution becomes impractical.
However, because the typical likelihood in SPR models describes the harvest numbers for a cohort as a multinomial distribution, we recommend using the Anscombe residuals for a binomial distribution (Anscombe 1953, Cox & Snell 1968) because the Pearson residuals can be skewed.
They can be modeled by a multinomial distribution with parameters and category response rates [p.
The information coming from length structures was evaluated modifying the sample size of the multinomial distribution in the likelihood function.
Use the multinomial distribution and the generalized logit link (available in SAS in the new Glimmix procedure).
j]|p, [mu], [theta], X) is a discrete multinomial distribution explicitly valued by