On Calibration Error of Randomized Forecasting Algorithms

V'yugin, Vladimir V.

doi:10.1007/978-3-540-75225-7_31

On Calibration Error of Randomized Forecasting Algorithms

Vladimir V. V'yugin

ALT 2007 pp. 388-402

doi:10.1007/978-3-540-75225-7_31 /alt/2007/vaposyugin2007alt-calibration/

Abstract

Recently, it was shown that calibration with an error less than δ > 0 is almost surely guaranteed with a randomized forecasting algorithm, where forecasts are chosen using randomized rounding up to δ of deterministic forecasts. We show that this error can not be improved for a large majority of sequences generated by a probabilistic algorithm: we prove that combining outcomes of coin-tossing and a transducer algorithm, it is possible to effectively generate with probability close to one a sequence “resistant” to any randomized rounding forecasting with an error much smaller than δ .

PDF ALT Semantic Scholar

Cite

Text

V'yugin. "On Calibration Error of Randomized Forecasting Algorithms." International Conference on Algorithmic Learning Theory, 2007. doi:10.1007/978-3-540-75225-7_31

Markdown

[V'yugin. "On Calibration Error of Randomized Forecasting Algorithms." International Conference on Algorithmic Learning Theory, 2007.](https://mlanthology.org/alt/2007/vaposyugin2007alt-calibration/) doi:10.1007/978-3-540-75225-7_31

BibTeX

@inproceedings{vaposyugin2007alt-calibration,
  title     = {{On Calibration Error of Randomized Forecasting Algorithms}},
  author    = {V'yugin, Vladimir V.},
  booktitle = {International Conference on Algorithmic Learning Theory},
  year      = {2007},
  pages     = {388-402},
  doi       = {10.1007/978-3-540-75225-7_31},
  url       = {https://mlanthology.org/alt/2007/vaposyugin2007alt-calibration/}
}