nach oben

European Actuarial Journal

Erschienen in:

11.01.2023 | Letter

Model selection with Gini indices under auto-calibration

verfasst von: Mario V. Wüthrich

Erschienen in: European Actuarial Journal | Ausgabe 1/2023

Einloggen

Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.

search-config

KI-gestützte Suche

Aus

Abstract

The Gini index does not give a strictly consistent scoring function. Therefore, simply maximizing the Gini index may lead to a wrong model choice. The main issue is that the Gini index is a rank-based score that is not calibration-sensitive. We show that the Gini index allows for strictly consistent scoring if we restrict it to the class of auto-calibrated regression models. That is, on the class of auto-calibrated models we know that the true model maximizes the Gini index.

Vorheriger Artikel A public micro pension programme in Brazil: heterogeneity among states and setting up of a benefit age adjustment

Nächster Artikel Does autocalibration improve goodness of lift?

Sie haben noch keine Lizenz? Dann Informieren Sie sich jetzt über unsere Produkte:

Springer Professional "Wirtschaft+Technik"

Online-Abonnement

Mit Springer Professional "Wirtschaft+Technik" erhalten Sie Zugriff auf:

über 102.000 Bücher
über 537 Zeitschriften

aus folgenden Fachgebieten:

Automobil + Motoren
Bauwesen + Immobilien
Business IT + Informatik
Elektrotechnik + Elektronik
Energie + Nachhaltigkeit
Finance + Banking
Management + Führung
Marketing + Vertrieb
Maschinenbau + Werkstoffe
Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

Jetzt informieren

Springer Professional "Wirtschaft"

Online-Abonnement

Mit Springer Professional "Wirtschaft" erhalten Sie Zugriff auf:

über 67.000 Bücher
über 340 Zeitschriften

aus folgenden Fachgebieten:

Bauwesen + Immobilien
Business IT + Informatik
Finance + Banking
Management + Führung
Marketing + Vertrieb
Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

Jetzt informieren

The continuity condition is a comparably weak assumption, e.g., it suffices to have one continuous covariate component and a strictly monotone regression function in that component; see Section 2.2 of Denuit et al. [4].

In (2.2) we assume that Y has a continuous distribution \(F_Y\), otherwise the denominator in (2.2) needs to be replaced by the term \({\mathbb {E}}[|Y-Z|]/(4{\mathbb {E}}[Y])\), where Z is an independent copy of Y, see also (4.1), below. For a continuous distribution \(F_Y\) the two terms in the denominator are then equivalent.

Byrne S (2016) A note on the use of empirical AUC for evaluating probabilistic forecasts. Electron J Stat 10:380–393MathSciNetCrossRefMATH

Ciatto N, Verelst H, Trufin J, Denuit M (2022) Does autocalibration improve goodness of lift? Eur Actuar J, in press

Denuit M, Charpentier A, Trufin J (2021) Autocalibration and Tweedie-dominance for insurance pricing in machine learning. Insurance 101:485–497MathSciNetMATH

Denuit M, Sznajder D, Trufin J (2019) Model selection based on Lorenz and concentration curves, Gini indices and convex order. Insurance 89:128–139MathSciNetMATH

Denuit M, Trufin J (2021) Lorenz curve, Gini coefficient, and Tweedie dominance for autocalibrated predictors. LIDAM Discussion Paper ISBA 2021/36

Dimitriadis T, Fissler T, Ziegel JF (2020) The efficiency gap. arXiv, 2010.14146

Engelmann B, Hayden E, Tasche D (2003) Testing rating accuracy. Risk 16(1):82–86

Frees EW, Meyers G, Cummings AD (2011) Summarizing insurance scores using a Gini index. J Am Stat Assoc 106:1085–1098MathSciNetCrossRefMATH

Frees EW, Meyers G, Cummings AD (2013) Insurance ratemaking and a Gini index. J Risk Insur 81:335–366CrossRef

10.

Gini C (1912) Variabilità e Mutuabilità. Contributo allo Studio delle Distribuzioni e delle Relazioni Statistiche. C, Cuppini, Bologna

11.

Gneiting T (2011) Making and evaluating point forecasts. J Am Stat Assoc 106(494):746–762MathSciNetCrossRefMATH

12.

Krüger F, Ziegel JF (2021) Generic conditions for forecast dominance. J Bus Econ Stat 39(4):972–983MathSciNetCrossRef

13.

Lorenz MO (1905) Methods of measuring the concentration of wealth. Publ Am Stat Assoc 9(70):209–219

14.

Menon AK, Jiang X, Vembu S, Elkan C, Ohno-Machado L (2012) Predicting accurate probabilities with ranking loss. ICML’12: Proceedings of the 29th International Conference on Machine Learning, 659–666

15.

Newson R (2002) Parameters behind “nonparametric’’ statistics: Kendall’s tau, Somers’ D and median differences. Stata J 2(1):45–64CrossRef

16.

Schervish MJ (1989) A general method of comparing probability assessors. Ann Stat 17(4):1856–1879MathSciNetCrossRefMATH

17.

So B, Boucher J, Valdez E (2021) Cost-sensitive multi-class AdaBoost for understanding driving behavior based on telematics. ASTIN Bull 51(3):719–751MathSciNetCrossRefMATH

18.

Somers RH (1962) A new asymmetric measure of association for ordinal variables. Am Sociol Rev 27(6):799–811CrossRef

19.

Tasche D (2006) Validation of internal rating systems and PD estimates. arXiv:0606071

20.

Tasche D (2021) Calibrating sufficiently. Statistics 55:1356–1386MathSciNetCrossRefMATH

21.

Tibshirani RJ, Hoefling H, Tibshirani R (2011) Nearly-isotonic regression. Technometrics 53(1):54–56MathSciNetCrossRef

22.

Wüthrich MV, Merz M (2023) Statistical foundations of actuarial learning and its applications. Springer Actuarial

Titel: Model selection with Gini indices under auto-calibration
verfasst von: Mario V. Wüthrich
Publikationsdatum: 11.01.2023
Verlag: Springer Berlin Heidelberg
Erschienen in: European Actuarial Journal / Ausgabe 1/2023
Print ISSN: 2190-9733
Elektronische ISSN: 2190-9741
DOI: https://doi.org/10.1007/s13385-022-00339-9

Springer Professional

Abstract

Bitte loggen Sie sich ein, um Zugang zu Ihrer Lizenz zu erhalten.

Sie haben noch keine Lizenz? Dann Informieren Sie sich jetzt über unsere Produkte:

Springer Professional "Wirtschaft+Technik"

Springer Professional "Wirtschaft"

Weitere Artikel der Ausgabe 1/2023

On some effects of dependencies on an insurer’s risk exposure, probability of ruin, and optimal premium loading

Smooth projection of mortality improvement rates: a Bayesian two-dimensional spline approach

Micro-level prediction of outstanding claim counts based on novel mixture models and neural networks

Does autocalibration improve goodness of lift?

Selection effect modification to the Lee-Carter model

Book review: pricing insurance risk – theory and practice