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

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Published

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

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Dzhamtyrova, R & Kalnishkan, Y 2019, Competitive Online Regression under Continuous Ranked Probability Score. in *Conformal and Probabilistic Prediction and Applications, 9-11 September 2019, Golden Sands, Bulgaria.* vol. 105, Proceedings of Machine Learning Research, Proceedings of Machine Learning Research, pp. 178-195, The 8th Symposium on Conformal and Probabilistic Prediction with Applications: COPA 2019, Golden Sands, Bulgaria, 9/09/19.

Dzhamtyrova, R., & Kalnishkan, Y. (2019). Competitive Online Regression under Continuous Ranked Probability Score. In *Conformal and Probabilistic Prediction and Applications, 9-11 September 2019, Golden Sands, Bulgaria *(Vol. 105, pp. 178-195). (Proceedings of Machine Learning Research). Proceedings of Machine Learning Research.

Dzhamtyrova R, Kalnishkan Y. Competitive Online Regression under Continuous Ranked Probability Score. In 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).

@inproceedings{4f32e04f839d4ab88932626bfaea6e1f,

title = "Competitive Online Regression under Continuous Ranked Probability Score",

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.",

author = "Raisa Dzhamtyrova and Yuri Kalnishkan",

year = "2019",

month = "9",

language = "English",

volume = "105",

series = "Proceedings of Machine Learning Research",

publisher = "Proceedings of Machine Learning Research",

pages = "178--195",

booktitle = "Conformal and Probabilistic Prediction and Applications, 9-11 September 2019, Golden Sands, Bulgaria",

}

TY - GEN

T1 - Competitive Online Regression under Continuous Ranked Probability Score

AU - Dzhamtyrova, Raisa

AU - Kalnishkan, Yuri

PY - 2019/9

Y1 - 2019/9

N2 - 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.

AB - 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.

M3 - Conference contribution

VL - 105

T3 - Proceedings of Machine Learning Research

SP - 178

EP - 195

BT - Conformal and Probabilistic Prediction and Applications, 9-11 September 2019, Golden Sands, Bulgaria

PB - Proceedings of Machine Learning Research

ER -