A Characterization of the Dirichlet Distribution with Application to Learning Bayesian Networks

UAI 1995 pp. 196-207

doi:10.1007/978-94-011-5430-7_7 /uai/1995/geiger1995uai-characterization/

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.

PDF UAI Semantic Scholar

Cite

Text

Geiger and Heckerman. "A Characterization of the Dirichlet Distribution with Application to Learning Bayesian Networks." Conference on Uncertainty in Artificial Intelligence, 1995. doi:10.1007/978-94-011-5430-7_7

Markdown

[Geiger and Heckerman. "A Characterization of the Dirichlet Distribution with Application to Learning Bayesian Networks." Conference on Uncertainty in Artificial Intelligence, 1995.](https://mlanthology.org/uai/1995/geiger1995uai-characterization/) doi:10.1007/978-94-011-5430-7_7

BibTeX

@inproceedings{geiger1995uai-characterization,
  title     = {{A Characterization of the Dirichlet Distribution with Application to Learning Bayesian Networks}},
  author    = {Geiger, Dan and Heckerman, David},
  booktitle = {Conference on Uncertainty in Artificial Intelligence},
  year      = {1995},
  pages     = {196-207},
  doi       = {10.1007/978-94-011-5430-7_7},
  url       = {https://mlanthology.org/uai/1995/geiger1995uai-characterization/}
}