### Abstract

We consider online prediction problems of combinatorial concepts. Examples of such concepts include s-t paths, permutations, truth assignments, set covers, and so on. The goal of the online prediction algorithm is to compete with the best fixed combinatorial concept in hindsight. A generic approach to this problem is to design an online prediction algorithm using the corresponding offline (approximation) algorithm as an oracle. The current state-of-the art method, however, is not efficient enough. In this paper we propose a more efficient online prediction algorithm when the offline approximation algorithm has a guarantee of the integrality gap.

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

Title of host publication | Algorithmic Learning Theory - 24th International Conference, ALT 2013, Proceedings |

Pages | 68-82 |

Number of pages | 15 |

DOIs | |

Publication status | Published - Nov 18 2013 |

Event | 24th International Conference on Algorithmic Learning Theory, ALT 2013 - Singapore, Singapore Duration: Oct 6 2013 → Oct 9 2013 |

### Publication series

Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
---|---|

Volume | 8139 LNAI |

ISSN (Print) | 0302-9743 |

ISSN (Electronic) | 1611-3349 |

### Other

Other | 24th International Conference on Algorithmic Learning Theory, ALT 2013 |
---|---|

Country | Singapore |

City | Singapore |

Period | 10/6/13 → 10/9/13 |

### Fingerprint

### All Science Journal Classification (ASJC) codes

- Theoretical Computer Science
- Computer Science(all)

### Cite this

*Algorithmic Learning Theory - 24th International Conference, ALT 2013, Proceedings*(pp. 68-82). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8139 LNAI). https://doi.org/10.1007/978-3-642-40935-6_6

**Combinatorial online prediction via metarounding.** / Fujita, Takahiro; Hatano, Kohei; Takimoto, Eiji.

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

*Algorithmic Learning Theory - 24th International Conference, ALT 2013, Proceedings.*Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 8139 LNAI, pp. 68-82, 24th International Conference on Algorithmic Learning Theory, ALT 2013, Singapore, Singapore, 10/6/13. https://doi.org/10.1007/978-3-642-40935-6_6

}

TY - GEN

T1 - Combinatorial online prediction via metarounding

AU - Fujita, Takahiro

AU - Hatano, Kohei

AU - Takimoto, Eiji

PY - 2013/11/18

Y1 - 2013/11/18

N2 - We consider online prediction problems of combinatorial concepts. Examples of such concepts include s-t paths, permutations, truth assignments, set covers, and so on. The goal of the online prediction algorithm is to compete with the best fixed combinatorial concept in hindsight. A generic approach to this problem is to design an online prediction algorithm using the corresponding offline (approximation) algorithm as an oracle. The current state-of-the art method, however, is not efficient enough. In this paper we propose a more efficient online prediction algorithm when the offline approximation algorithm has a guarantee of the integrality gap.

AB - We consider online prediction problems of combinatorial concepts. Examples of such concepts include s-t paths, permutations, truth assignments, set covers, and so on. The goal of the online prediction algorithm is to compete with the best fixed combinatorial concept in hindsight. A generic approach to this problem is to design an online prediction algorithm using the corresponding offline (approximation) algorithm as an oracle. The current state-of-the art method, however, is not efficient enough. In this paper we propose a more efficient online prediction algorithm when the offline approximation algorithm has a guarantee of the integrality gap.

UR - http://www.scopus.com/inward/record.url?scp=84887443709&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84887443709&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-40935-6_6

DO - 10.1007/978-3-642-40935-6_6

M3 - Conference contribution

AN - SCOPUS:84887443709

SN - 9783642409349

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 68

EP - 82

BT - Algorithmic Learning Theory - 24th International Conference, ALT 2013, Proceedings

ER -