Combinatorial Online Prediction via Metarounding

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.

Cite

Text

Fujita et al. "Combinatorial Online Prediction via Metarounding." International Conference on Algorithmic Learning Theory, 2013. doi:10.1007/978-3-642-40935-6_6

Markdown

[Fujita et al. "Combinatorial Online Prediction via Metarounding." International Conference on Algorithmic Learning Theory, 2013.](https://mlanthology.org/alt/2013/fujita2013alt-combinatorial/) doi:10.1007/978-3-642-40935-6_6

BibTeX

@inproceedings{fujita2013alt-combinatorial,
  title     = {{Combinatorial Online Prediction via Metarounding}},
  author    = {Fujita, Takahiro and Hatano, Kohei and Takimoto, Eiji},
  booktitle = {International Conference on Algorithmic Learning Theory},
  year      = {2013},
  pages     = {68-82},
  doi       = {10.1007/978-3-642-40935-6_6},
  url       = {https://mlanthology.org/alt/2013/fujita2013alt-combinatorial/}
}