A Characterization of the Dirichlet Distribution with Application to Learning Bayesian Networks
arXiv:1302.4949
Abstract
We provide a new characterization of the Dirichlet distribution. This characterization implies that under assumptions made by several previous authors for learning belief networks, a Dirichlet prior on the parameters is inevitable.
Appears in Proceedings of the Eleventh Conference on Uncertainty in Artificial Intelligence (UAI1995)