Elicitation and Identification of Properties
Abstract
Properties of distributions are real-valued functionals such as the mean, quantile or conditional value at risk. A property is elicitable if there exists a scoring function such that minimization of the associated risks recovers the property. We extend existing results to characterize the elicitability of properties in a general setting. We further relate elicitability to identifiability (a notion introduced by Osband) and provide a general formula describing all scoring functions for an elicitable property. Finally, we draw some connections to the theory of coherent risk measures.
Cite
Text
Steinwart et al. "Elicitation and Identification of Properties." Annual Conference on Computational Learning Theory, 2014.Markdown
[Steinwart et al. "Elicitation and Identification of Properties." Annual Conference on Computational Learning Theory, 2014.](https://mlanthology.org/colt/2014/steinwart2014colt-elicitation/)BibTeX
@inproceedings{steinwart2014colt-elicitation,
title = {{Elicitation and Identification of Properties}},
author = {Steinwart, Ingo and Pasin, Chloé and Williamson, Robert C. and Zhang, Siyu},
booktitle = {Annual Conference on Computational Learning Theory},
year = {2014},
pages = {482-526},
url = {https://mlanthology.org/colt/2014/steinwart2014colt-elicitation/}
}