(2000; revised 2003, 2009, 2012)

The Dirichlet distribution and its compound variant, the Dirichlet-multinomial, are two of the most basic models for proportional data, such as the mix of vocabulary words in a text document. Yet the maximum-likelihood estimate of these distributions is not available in closed-form. This paper describes simple and efficient iterative schemes for obtaining parameter estimates in these models. In each case, a fixed-point iteration and a Newton-Raphson (or generalized Newton-Raphson) iteration is provided.

Matlab functions which implement these algorithms are available in fastfit.

PDF (134K)

Thomas P Minka Last modified: Wed Aug 17 18:10:52 GMT 2005