### Abstract

This chapter addresses the problem of estimating the parameters of a Bayesian network from incomplete data. This is a hard problem, which for computational reasons cannot be effectively tackled by a full Bayesian approach. The work around is to search for the estimate with maximum posterior probability. This is usually done by selecting the highest posterior probability estimate among those found by multiple runs of Expectation-Maximization with distinct starting points. However, many local maxima characterize the posterior probability function, and several of them have similar high probability. We argue that high probability is necessary but not sufficient in order to obtain good estimates.We present an approach based on maximum entropy to address this problem and describe a simple and effective way to implement it. Experiments show that our approach produces significantly better estimates than the most commonly used method.

Original language | English |
---|---|

Title of host publication | Interdisciplinary Bayesian Statistics, EBEB 2014 |

Editors | Adriano Polpo, Francisco Louzada, Marcelo Lauretto, Julio Michael Stern, Laura Letícia Ramos Rifo |

Publisher | Springer |

Chapter | 6 |

Pages | 69-82 |

Number of pages | 14 |

Volume | 118 |

ISBN (Electronic) | 9783319124537 |

DOIs | |

Publication status | Published - 1 Jan 2015 |

Externally published | Yes |

Event | 12th Brazilian Meeting on Bayesian Statistics, EBEB 2014 - Atibaia, Brazil Duration: 10 Mar 2014 → 14 Mar 2014 |

### Conference

Conference | 12th Brazilian Meeting on Bayesian Statistics, EBEB 2014 |
---|---|

Country | Brazil |

City | Atibaia |

Period | 10/03/14 → 14/03/14 |

## Fingerprint Dive into the research topics of 'A maximum entropy approach to learn Bayesian networks from incomplete data'. Together they form a unique fingerprint.

## Cite this

*Interdisciplinary Bayesian Statistics, EBEB 2014*(Vol. 118, pp. 69-82). Springer. https://doi.org/10.1007/978-3-319-12454-4_6