Competitive Online Regression under Continuous Ranked Probability Score. / Dzhamtyrova, Raisa; Kalnishkan, Yuri.

Conformal and Probabilistic Prediction and Applications, 9-11 September 2019, Golden Sands, Bulgaria. Vol. 105 Proceedings of Machine Learning Research, 2019. p. 178-195 (Proceedings of Machine Learning Research).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Published

Abstract

We consider the framework of competitive prediction when one provides guarantees compared to other predictive models that are called experts. We propose the algorithm that combines point predictions of an infinite pool of linear experts and outputs probability forecasts in the form of cumulative distribution functions. We evaluate the quality of probabilistic prediction by the continuous ranked probability score (CRPS), which is a widely used proper scoring rule. We provide a strategy that allows us to 'track the best expert' and derive the theoretical bound on the discounted loss of the strategy. Experimental results on synthetic data and solar power data show that the theoretical bounds of our algorithm are not violated. Also the algorithm performs close to and sometimes outperforms the retrospectively best quantile regression.
Original languageEnglish
Title of host publicationConformal and Probabilistic Prediction and Applications, 9-11 September 2019, Golden Sands, Bulgaria
PublisherProceedings of Machine Learning Research
Pages178-195
Number of pages18
Volume105
Publication statusPublished - Sep 2019
EventThe 8th Symposium on Conformal and Probabilistic Prediction with Applications: COPA 2019 - Golden Sands, Bulgaria
Duration: 9 Sep 201911 Sep 2019
http://clrc.rhul.ac.uk/copa2019/

Publication series

NameProceedings of Machine Learning Research
ISSN (Electronic)2640-3498

Conference

ConferenceThe 8th Symposium on Conformal and Probabilistic Prediction with Applications: COPA 2019
CountryBulgaria
CityGolden Sands
Period9/09/1911/09/19
Internet address
This open access research output is licenced under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.

ID: 34371029