Supermartingales in prediction with expert advice
| dc.contributor.author | Chernov, Alexey | en_GB |
| dc.contributor.author | Kalnishkan, Yuri | en_GB |
| dc.contributor.author | Zhdanov, Fedor | en_GB |
| dc.contributor.author | Vovk, Vladimir | en_GB |
| dc.date.accessioned | 2013-04-07T16:46:49Z | |
| dc.date.available | 2013-04-07T16:46:49Z | |
| dc.date.issued | 2010 | |
| dc.identifier.citation | Chernov, A., Kalnishkan, Y., Zhdanov, F. and Vovk, V., (2010) 'Supermartingales in prediction with expert advice' 411 (29-30):2647-2669 Theoretical Computer Science | en_GB |
| dc.identifier.issn | 0304-3975 | |
| dc.identifier.doi | 10.1016/j.tcs.2010.04.003 | |
| dc.identifier.uri | http://hdl.handle.net/10547/279178 | |
| dc.description.abstract | The paper applies the method of defensive forecasting, based on the use of game-theoretic supermartingales, to prediction with expert advice. In the traditional setting of a countable number of experts and a finite number of outcomes, the Defensive Forecasting Algorithm is very close to the well-known Aggregating Algorithm. Not only the performance guarantees but also the predictions are the same for these two methods of fundamentally different nature. The paper also discusses a new setting where the experts can give advice conditional on the learner's future decision. Both the algorithms can be adapted to the new setting and give the same performance guarantees as in the traditional setting. Finally, an application of defensive forecasting to a setting with several loss functions is outlined. | |
| dc.language.iso | en | en |
| dc.publisher | Elsevier | en_GB |
| dc.relation.url | http://linkinghub.elsevier.com/retrieve/pii/S0304397510001982 | en_GB |
| dc.rights | Archived with thanks to Theoretical Computer Science | en_GB |
| dc.title | Supermartingales in prediction with expert advice | en |
| dc.type | Article | en |
| dc.identifier.journal | Theoretical Computer Science | en_GB |
| html.description.abstract | The paper applies the method of defensive forecasting, based on the use of game-theoretic supermartingales, to prediction with expert advice. In the traditional setting of a countable number of experts and a finite number of outcomes, the Defensive Forecasting Algorithm is very close to the well-known Aggregating Algorithm. Not only the performance guarantees but also the predictions are the same for these two methods of fundamentally different nature. The paper also discusses a new setting where the experts can give advice conditional on the learner's future decision. Both the algorithms can be adapted to the new setting and give the same performance guarantees as in the traditional setting. Finally, an application of defensive forecasting to a setting with several loss functions is outlined. |
Files in this item
Name:
Publisher version