### Resumé

A class of probabilistic-logic models is considered, which increases the expressibility from HMM's and SCFG's regular and context-free languages to, in principle, Turing complete languages. In general, such models are computationally far too complex for direct use, so optimization by pruning and approximation are needed. The first steps are taken towards a methodology for optimizing such models by approximations using auxiliary models for preprocessing or splitting them into submodels. Evaluation of such approximating models is challenging as authoritative test data may be sparse. On the other hand, the original complex models may be used for generating artificial evaluation data by efficient sampling, which can be used in the evaluation, although it does not constitute a foolproof test procedure. These models and evaluation processes are illustrated in the PRISM system developed by other authors, and we discuss their applicability and limitations.

Originalsprog | Engelsk |
---|---|

Bogserie | Lecture Notes in Computer Science |

Sider (fra-til) | 70-83 |

Antal sider | 14 |

ISSN | 0302-9743 |

Status | Udgivet - 2009 |

Begivenhed | International Conference on Logic Programming 2009 - Pasadena, USA Varighed: 11 jul. 2009 → 17 jul. 2009 Konferencens nummer: 25 |

### Konference

Konference | International Conference on Logic Programming 2009 |
---|---|

Nummer | 25 |

Land | USA |

By | Pasadena |

Periode | 11/07/2009 → 17/07/2009 |

### Bibliografisk note

Volumne: 5649### Citer dette

}

**Preprocessing for Optimization of Probabilistic-Logic Models for Sequence Analysis.** / Christiansen, Henning; Lassen, Ole Torp.

Publikation: Bidrag til tidsskrift › Konferenceartikel › Forskning › peer review

TY - GEN

T1 - Preprocessing for Optimization of Probabilistic-Logic Models for Sequence Analysis

AU - Christiansen, Henning

AU - Lassen, Ole Torp

N1 - Volumne: 5649

PY - 2009

Y1 - 2009

N2 - A class of probabilistic-logic models is considered, which increases the expressibility from HMM's and SCFG's regular and context-free languages to, in principle, Turing complete languages. In general, such models are computationally far too complex for direct use, so optimization by pruning and approximation are needed. The first steps are taken towards a methodology for optimizing such models by approximations using auxiliary models for preprocessing or splitting them into submodels. Evaluation of such approximating models is challenging as authoritative test data may be sparse. On the other hand, the original complex models may be used for generating artificial evaluation data by efficient sampling, which can be used in the evaluation, although it does not constitute a foolproof test procedure. These models and evaluation processes are illustrated in the PRISM system developed by other authors, and we discuss their applicability and limitations.

AB - A class of probabilistic-logic models is considered, which increases the expressibility from HMM's and SCFG's regular and context-free languages to, in principle, Turing complete languages. In general, such models are computationally far too complex for direct use, so optimization by pruning and approximation are needed. The first steps are taken towards a methodology for optimizing such models by approximations using auxiliary models for preprocessing or splitting them into submodels. Evaluation of such approximating models is challenging as authoritative test data may be sparse. On the other hand, the original complex models may be used for generating artificial evaluation data by efficient sampling, which can be used in the evaluation, although it does not constitute a foolproof test procedure. These models and evaluation processes are illustrated in the PRISM system developed by other authors, and we discuss their applicability and limitations.

KW - Logic programming

KW - bio-informatics

KW - probabilistic sequence analysis

M3 - Conference article

SP - 70

EP - 83

JO - Lecture Notes in Computer Science

JF - Lecture Notes in Computer Science

SN - 0302-9743

ER -