### Abstract

Modeling of a large joint probability table is problematic when its variables have a large number of categories. In such a case, a mixture of simpler probability tables could be a good model. And the estimation of such a large probability table frequently has another problem of data sparseness. When constructing mixture models with sparse data, EM estimators based on the β-likelihood are expected more appropriate than those based on the log likelihood. Experimental results show that a mixture model estimated by the βlikelihood approximates a large joint probability table with sparse data more appropriately than EM estimators.

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

Title of host publication | Proceedings of the Sixth SIAM International Conference on Data Mining |

Pages | 519-523 |

Number of pages | 5 |

Volume | 2006 |

Publication status | Published - 2006 |

Event | Sixth SIAM International Conference on Data Mining - Bethesda, MD Duration: 2006 Apr 20 → 2006 Apr 22 |

### Other

Other | Sixth SIAM International Conference on Data Mining |
---|---|

City | Bethesda, MD |

Period | 06/4/20 → 06/4/22 |

### Keywords

- β-likelihood
- EM estimation
- Mixture model
- Robustness
- Sparse data

### ASJC Scopus subject areas

- Engineering(all)

### Cite this

*Proceedings of the Sixth SIAM International Conference on Data Mining*(Vol. 2006, pp. 519-523)

**Robust estimation for mixture of probability tables based on β-likelihood.** / Fujimoto, Yu; Murata, Noboru.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

*Proceedings of the Sixth SIAM International Conference on Data Mining.*vol. 2006, pp. 519-523, Sixth SIAM International Conference on Data Mining, Bethesda, MD, 06/4/20.

}

TY - GEN

T1 - Robust estimation for mixture of probability tables based on β-likelihood

AU - Fujimoto, Yu

AU - Murata, Noboru

PY - 2006

Y1 - 2006

N2 - Modeling of a large joint probability table is problematic when its variables have a large number of categories. In such a case, a mixture of simpler probability tables could be a good model. And the estimation of such a large probability table frequently has another problem of data sparseness. When constructing mixture models with sparse data, EM estimators based on the β-likelihood are expected more appropriate than those based on the log likelihood. Experimental results show that a mixture model estimated by the βlikelihood approximates a large joint probability table with sparse data more appropriately than EM estimators.

AB - Modeling of a large joint probability table is problematic when its variables have a large number of categories. In such a case, a mixture of simpler probability tables could be a good model. And the estimation of such a large probability table frequently has another problem of data sparseness. When constructing mixture models with sparse data, EM estimators based on the β-likelihood are expected more appropriate than those based on the log likelihood. Experimental results show that a mixture model estimated by the βlikelihood approximates a large joint probability table with sparse data more appropriately than EM estimators.

KW - β-likelihood

KW - EM estimation

KW - Mixture model

KW - Robustness

KW - Sparse data

UR - http://www.scopus.com/inward/record.url?scp=33745432710&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33745432710&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:33745432710

SN - 089871611X

SN - 9780898716115

VL - 2006

SP - 519

EP - 523

BT - Proceedings of the Sixth SIAM International Conference on Data Mining

ER -