Elicitation and Identification of Properties

Ingo Steinwart, Chloé Pasin, Robert Williamson, Siyu Zhang
Proceedings of The 27th Conference on Learning Theory, PMLR 35:482-526, 2014.

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 this Paper

BibTeX
@InProceedings{pmlr-v35-steinwart14, title = {Elicitation and Identification of Properties}, author = {Steinwart, Ingo and Pasin, Chloé and Williamson, Robert and Zhang, Siyu}, booktitle = {Proceedings of The 27th Conference on Learning Theory}, pages = {482--526}, year = {2014}, editor = {Balcan, Maria Florina and Feldman, Vitaly and Szepesvári, Csaba}, volume = {35}, series = {Proceedings of Machine Learning Research}, address = {Barcelona, Spain}, month = {13--15 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v35/steinwart14.pdf}, url = {https://proceedings.mlr.press/v35/steinwart14.html}, 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.} }
Endnote
%0 Conference Paper %T Elicitation and Identification of Properties %A Ingo Steinwart %A Chloé Pasin %A Robert Williamson %A Siyu Zhang %B Proceedings of The 27th Conference on Learning Theory %C Proceedings of Machine Learning Research %D 2014 %E Maria Florina Balcan %E Vitaly Feldman %E Csaba Szepesvári %F pmlr-v35-steinwart14 %I PMLR %P 482--526 %U https://proceedings.mlr.press/v35/steinwart14.html %V 35 %X 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.
RIS
TY - CPAPER TI - Elicitation and Identification of Properties AU - Ingo Steinwart AU - Chloé Pasin AU - Robert Williamson AU - Siyu Zhang BT - Proceedings of The 27th Conference on Learning Theory DA - 2014/05/29 ED - Maria Florina Balcan ED - Vitaly Feldman ED - Csaba Szepesvári ID - pmlr-v35-steinwart14 PB - PMLR DP - Proceedings of Machine Learning Research VL - 35 SP - 482 EP - 526 L1 - http://proceedings.mlr.press/v35/steinwart14.pdf UR - https://proceedings.mlr.press/v35/steinwart14.html AB - 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. ER -
APA
Steinwart, I., Pasin, C., Williamson, R. & Zhang, S.. (2014). Elicitation and Identification of Properties. Proceedings of The 27th Conference on Learning Theory, in Proceedings of Machine Learning Research 35:482-526 Available from https://proceedings.mlr.press/v35/steinwart14.html.

Related Material