Contribution of expert aggregation to temperature prediction part II: Second order bounds with sleeping experts

TL;DR

This paper enhances temperature prediction accuracy by integrating sleeping expert frameworks with online expert aggregation, employing gradient boosted trees for adaptive, second-order regret bounds, and optimizing expert selection to reduce errors.

Contribution

It introduces a reactive aggregation method using the Sleeping Expert Framework combined with gradient boosted regression trees, improving temperature prediction accuracy and error bounds.

Findings

Reduced large errors in temperature predictions.

Achieved better responsiveness in expert aggregation.

Maintained or improved root mean squared error.

Abstract

In this paper we improve on the temperature predictions made with (online) Expert Aggregation (EA) [Cesa-Bianchi and Lugosi, 2006] in Part I. In particular, we make the aggregation more reactive, whilst maintaining at least the same root mean squared error and reducing the number of large errors. We have achieved this by using the Sleeping Expert Framework (SEF) [Freund et al., 1997, Devaine et al., 2013], which allows the more efficient use of biased experts (bad on average but which may be good at some point). To deal with the fact that, unlike in Devaine et al. [2013], we do not know in advance when to use these biased experts, we resorted to gradient boosted regression trees [Chen and Guestrin, 2016] and provide regret bounds against sequences of experts [Mourtada and Maillard, 2017] which take into account this uncertainty. We applied this in a fully online way on BOA [Wintenberger, 2024], an adaptive aggregation with second order regret bounds, which had the best results in Part I. Finally, we made a meta-aggregation with the EA follow the leader. This chooses whether or not to use the SEF in order to limit the possible noise added by the SEF.