Aggregating Algorithm for prediction of packs. / Adamskiy, Dmitry; Bellotti, Tony; Dzhamtyrova, Raisa; Kalnishkan, Yury.

In: Machine Learning, 07.01.2019, p. 1-30.

Research output: Contribution to journalArticle

E-pub ahead of print

Documents

  • Accepted Manuscript

    Accepted author manuscript, 971 KB, PDF-document

Links

Abstract

This paper formulates a protocol for prediction of packs, which is a special case of on-line prediction under delayed feedback. Under the prediction of packs protocol, the learner must make a few predictions without seeing the respective outcomes and then the outcomes are revealed in one go. The paper develops the theory of prediction with expert advice for packs by generalising the concept of mixability. We propose a number of merging algorithms for prediction of packs with tight worst case loss upper bounds similar to those for Vovk's Aggregating Algorithm. Unlike existing algorithms for delayed feedback settings, our algorithms do not depend on the order of outcomes in a pack. Empirical experiments on sports and house price datasets are carried out to study the performance of the new algorithms and compare them against an existing method.
Original languageEnglish
Pages (from-to)1-30
Number of pages30
JournalMachine Learning
Early online date7 Jan 2019
DOIs
StateE-pub ahead of print - 7 Jan 2019
This open access research output is licenced under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.

ID: 31973161