×
Giummolè, Federica; Mameli, Valentina

Asymptotic minimum scoring rule prediction. (English) Zbl 1395.62292

Electron. J. Stat. 12, No. 2, 2401-2429 (2018).
Summary: Most of the methods nowadays employed in forecast problems are based on scoring rules. There is a divergence function associated to each scoring rule, that can be used as a measure of discrepancy between probability distributions. This approach is commonly used in the literature for comparing two competing predictive distributions on the basis of their relative expected divergence from the true distribution.
In this paper we focus on the use of scoring rules as a tool for finding predictive distributions for an unknown of interest. The proposed predictive distributions are asymptotic modifications of the estimative solutions, obtained by minimizing the expected divergence related to a general scoring rule.
The asymptotic properties of such predictive distributions are strictly related to the geometry induced by the considered divergence on a regular parametric model. In particular, the existence of a global optimal predictive distribution is guaranteed for invariant divergences, whose local behaviour is similar to well known \(\alpha\)-divergences.
We show that a wide class of divergences obtained from weighted scoring rules share invariance properties with \(\alpha\)-divergences. For weighted scoring rules it is thus possible to obtain a global solution to the prediction problem. Unfortunately, the divergences associated to many widely used scoring rules are not invariant. Still for these cases we provide a locally optimal predictive distribution, within a specified parametric model.

MSC:

62M20 Inference from stochastic processes and prediction
60G25 Prediction theory (aspects of stochastic processes)

Keywords:

\(\alpha \)-connection; Fisher metric; Kullback-Leibler divergence; monotone and regular divergence; predictive density; scoring rule; weighted scoring rule
Cite Review PDF
Full Text: DOI Euclid

References:

[1] Aitchison, J. (1975). Goodness of prediction fit., Biometrika62 547–554. · Zbl 0339.62018 · doi:10.1093/biomet/62.3.547
[2] Aitchison, J. and Dunsmore, I.R. (1975)., Statistical prediction analysis. Cambridge University Press, Cambridge. · Zbl 0327.62043
[3] Amari, S. (1985)., Differential Geometric Methods in Statistics. Lecture Notes in Statistics, 28. New York: Springer-Verlag. · Zbl 0559.62001
[4] Amari, S. (2010). Divergence function, information monotonicity and information geometry., Bulletin of the Polish Academy of Sciences: Technical Sciences58 183–195.
[5] Barndorff-Nielsen, O.E. and Cox, D.R. (1996). Prediction and asymptotics., Bernoulli2 319–340. · Zbl 0870.62008 · doi:10.2307/3318417
[6] Bjørnstad, J.F. (1990). Predictive likelihood: A review., Statistical Sciences5 242–265. · Zbl 0955.62517
[7] Brier, G.W. (1950). Verification of forecasts expressed in terms of probability., Monthly Weather Review78 1–3.
[8] Ĉencov, N.N. (1982)., Statistical decision rules and optimal inference. Translations of Mathematical Monographs 53. AMS, Providence, RI. · Zbl 0484.62008
[9] Corcuera, J.M. and Giummolè, F. (1998). A characterization of monotone and regular divergences., Annals of the Institute of Statistical Mathematics50 433–450. · Zbl 0922.62008 · doi:10.1023/A:1003569210573
[10] Corcuera, J.M. and Giummolè, F. (1999). On the relationship between \(α \)-connections and the asymptotic properties of predictive distributions., Bernoulli5 163–176. · Zbl 0916.62014 · doi:10.2307/3318617
[11] Corcuera, J.M. and Giummolè, F. (2000). First order optimal predictive densities. In Marriott, P., Salmon, M., Applications of differential geometry to econometrics, Cambridge University Press.
[12] Csiszár, I. (1967). Information-type measure of difference of probability distributions and indirect observations., Studia Scientiarum Mathematicarum Hungarica2 299–318. · Zbl 0157.25802
[13] Dawid, A.P. (1998). Coherent measures of discrepancy, uncertainty and dependence, with applications to Bayesian predictive experimental design., Technical Report 139, Department of Statistical Science, University College London. http://www.ucl.ac.uk/Stats/research/pdfs/139b.zip
[14] Dawid, A.P. (2007). The geometry of proper scoring rules., Annals of the Institute of Statistical Mathematics59 77–93. · Zbl 1108.62009 · doi:10.1007/s10463-006-0099-8
[15] Dawid, A.P. and Musio, M. (2014). Theory and applications of proper scoring rules., Metron72 169–183. · Zbl 1316.62013 · doi:10.1007/s40300-014-0039-y
[16] Dawid, A.P., Musio, M. and Ventura, L. (2016). Minimum scoring rule inference., Scandinavian Journal of Statistics43 123–138. · Zbl 1371.62019 · doi:10.1111/sjos.12168
[17] Eguchi, S. (1992). Geometry of minimum contrast., Hiroshima Mathematical Journal22 631–647. · Zbl 0780.53015 · doi:10.32917/hmj/1206128508
[18] Eguchi, S. (2006). Information geometry and statistical pattern recognition., Sugaku Exposition. AMS, Providence, RI. · Zbl 1277.62164
[19] Forbes, P.G.M. (2012). Compatible weighted proper scoring rules., Biometrika99 989–994. · Zbl 1452.62153 · doi:10.1093/biomet/ass046
[20] Fonseca, G., Giummolè, F. and Vidoni, P. (2014). Calibrating predictive distributions., Journal of Statistical Computation and Simulation84 373–383. · Zbl 1459.62177
[21] Giummolè, F., Ventura, L. (2006). Robust prediction limits based on M-estimators., Statistics and Probability Letters76 1735–1740. · Zbl 1098.62034 · doi:10.1016/j.spl.2006.04.013
[22] Gneiting, T. and Raftery, A.E. (2007). Strictly proper scoring rules, prediction, and estimation., Journal of the American Statistical Association102 359–378. · Zbl 1284.62093 · doi:10.1198/016214506000001437
[23] Gneiting,T. and Katzfuss, M. (2014). Probabilistic Forecasting., Annual Review of Statistics and Its Application1 125–151.
[24] Good, I.J. (1952). Rational decisions., Journal of the Royal Statistical Society, Series B14 107–114.
[25] Hall, P., Peng, L. and Tajvidi, N. (1999). On prediction intervals based on predictive likelihood or bootstrap methods., Biometrika86 871–880. · Zbl 0942.62027 · doi:10.1093/biomet/86.4.871
[26] Harris, I.R. (1989). Predictive fit for natural exponential families., Biometrika76 675–684. · Zbl 0679.62021 · doi:10.1093/biomet/76.4.675
[27] Holzmann, H. and Klar, B. (2017). Focusing on regions of interest in forecast evaluation., The Annals of Applied Statistics11(4), 2404–2431. · Zbl 1383.62243 · doi:10.1214/17-AOAS1088
[28] Kass, S. and Vos, P.W. (1997)., Geometrical Foundations of Asymptotic Inference, Wiley Series in Probability and Statistics. New York: John Wiley & Sons, Inc. · Zbl 0880.62005
[29] Komaki, F. (1996). On asymptotic properties of predictive distributions., Biometrika83 299–313. · Zbl 0864.62007 · doi:10.1093/biomet/83.2.299
[30] Jose, V.R.R., Nau, R.F. and Winkler, R.L. (2009). Scoring rules, generalized entropy, and utility maximization., Operations Research56 1146–1157. · Zbl 1167.91389 · doi:10.1287/opre.1070.0498
[31] Jose, V.R.R. (2008)., The Verification of Probabilistic Forecasts in Decision and Risk Analysis. Phd Thesis, Department of Business Administration, Duke University.
[32] Lawless, J.F. and Fredette, M. (2005). Frequentist prediction intervals and predictive distributions., Biometrika92 529–542. · Zbl 1183.62052 · doi:10.1093/biomet/92.3.529
[33] Machete, R. (2013). Contrasting probabilistic scoring rules., Journal of Statistical Planning and Inference143 1781–1790. · Zbl 1279.62111 · doi:10.1016/j.jspi.2013.05.012
[34] Mameli, V. and Ventura, L. (2015). Higher-order asymptotics for scoring rules., Journal of Statistical Planning and Inference165 13–26. · Zbl 1326.62023 · doi:10.1016/j.jspi.2015.03.005
[35] Mameli, V., Musio, M. and Ventura, L. (2018). Bootstrap adjustments of signed scoring rule root statistics., Communications in Statistics – Simulation and Computation47 1204–1215. · Zbl 07549519
[36] Mendenez, M.L., Morales, D., Pardo, L. and Salicrù, M. (1997). \((h,ϕ )\)-entropy differential metric., Applications of Mathematics42 81–98. · Zbl 0898.62005
[37] Murray, G.D. (1977). A note on the estimation of probability density functions., Biometrika64 150–152. · Zbl 0347.62035 · doi:10.2307/2335788
[38] Murray, M.K. and Rice, J.W. (1993)., Differential Geometry and Statistics, Monographs on Statistics and Applied Probability, 48. London: Chapman & Hall. · Zbl 0804.53001
[39] Pardo, L. (2006)., Statistical inference based on divergence measure. Florida: Boca Raton, Taylor & Francis. · Zbl 1118.62008
[40] Savage, L.J. (1971). Elicitation of Personal Probabilities and Expectations., Journal of the American Statistical Association66 783–801. · Zbl 0253.92008 · doi:10.1080/01621459.1971.10482346
[41] Tsallis, C. (1988). Possible generalization of Boltzmann-Gibbs statistics., Journal of Statistical Physics52 479–487. · Zbl 1082.82501 · doi:10.1007/BF01016429
[42] Vidoni, P. (1995). A simple predictive density based on the \(p^*\)-formula., Biometrika82 855–863. · Zbl 0878.62017
[43] Vidoni, P. (2009). Improved prediction intervals and distribution functions., Scandinavian Journal of Statistics36 735–748. · Zbl 1222.62027 · doi:10.1111/j.1467-9469.2009.00656.x
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.