1
|
Tyralis H, Papacharalampous G, Dogulu N, Chun KP. Deep Huber quantile regression networks. Neural Netw 2025; 187:107364. [PMID: 40112635 DOI: 10.1016/j.neunet.2025.107364] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/09/2023] [Revised: 01/06/2025] [Accepted: 03/04/2025] [Indexed: 03/22/2025]
Abstract
Typical machine learning regression applications aim to report the mean or the median of the predictive probability distribution, via training with a squared or an absolute error scoring function. The importance of issuing predictions of more functionals of the predictive probability distribution (quantiles and expectiles) has been recognized as a means to quantify the uncertainty of the prediction. In deep learning (DL) applications, that is possible through quantile and expectile regression neural networks (QRNN and ERNN respectively). Here we introduce deep Huber quantile regression networks (DHQRN) that nest QRNN and ERNN as edge cases. DHQRN can predict Huber quantiles, which are more general functionals in the sense that they nest quantiles and expectiles as limiting cases. The main idea is to train a DL algorithm with the Huber quantile scoring function, which is consistent for the Huber quantile functional. As a proof of concept, DHQRN are applied to predict house prices in Melbourne, Australia and Boston, United States (US). In this context, predictive performances of three DL architectures are discussed along with evidential interpretation of results from two economic case studies. Additional simulation experiments and applications to real-world case studies using open datasets demonstrate a satisfactory absolute performance of DHQRN.
Collapse
Affiliation(s)
- Hristos Tyralis
- Department of Topography, School of Rural, Surveying and Geoinformatics Engineering, National Technical University of Athens, Iroon Polytechniou 5, Zografou 157 80, Greece; Construction Agency, Hellenic Air Force, Mesogion Avenue 227-231, Cholargos 15 561, Greece.
| | - Georgia Papacharalampous
- Department of Topography, School of Rural, Surveying and Geoinformatics Engineering, National Technical University of Athens, Iroon Polytechniou 5, Zografou 157 80, Greece
| | - Nilay Dogulu
- Hydrology, Water Resources and Cryosphere Branch, World Meteorological Organisation (WMO), Geneva, Switzerland
| | - Kwok P Chun
- Department of Geography and Environmental Management, University of the West of England, Bristol, United Kingdom
| |
Collapse
|
2
|
Allouche M, Girard S, Gobet E. Learning extreme expected shortfall and conditional tail moments with neural networks. Application to cryptocurrency data. Neural Netw 2025; 182:106903. [PMID: 39608147 DOI: 10.1016/j.neunet.2024.106903] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/03/2024] [Revised: 10/28/2024] [Accepted: 11/08/2024] [Indexed: 11/30/2024]
Abstract
We propose a neural networks method to estimate extreme Expected Shortfall, and even more generally, extreme conditional tail moments as functions of confidence levels, in heavy-tailed settings. The convergence rate of the uniform error between the log-conditional tail moment and its neural network approximation is established leveraging extreme-value theory (in particular the high-order condition on the distribution tails) and using critically two activation functions (eLU and ReLU) for neural networks. The finite sample performance of the neural network estimator is compared to bias-reduced extreme-value competitors using synthetic heavy-tailed data. The experiments reveal that our method largely outperforms others. In addition, the selection of the anchor point appears to be much easier and stabler than for other methods. Finally, the neural network estimator is tested on real data related to extreme loss returns in cryptocurrencies: here again, the accuracy obtained by cross-validation is excellent, and is much better compared with competitors.
Collapse
Affiliation(s)
- Michaël Allouche
- Kaiko - Quantitative Data, 2 rue de Choiseul, Paris, 75002, France
| | - Stéphane Girard
- Univ. Grenoble Alpes, Inria, CNRS, Grenoble INP, LJK, Grenoble, 38000, France.
| | - Emmanuel Gobet
- Centre de Mathématiques Appliquées (CMAP), CNRS, Ecole Polytechnique, Institut Polytechnique de Paris, Route de Saclay, Palaiseau, 91128, France
| |
Collapse
|
3
|
Arbel J, Girard S, Nguyen HD, Usseglio-Carleve A. Multivariate expectile-based distribution: Properties, Bayesian inference, and applications. J Stat Plan Inference 2023. [DOI: 10.1016/j.jspi.2022.12.001] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 12/23/2022]
|
4
|
Local linear estimate of the functional expectile regression. Stat Probab Lett 2023. [DOI: 10.1016/j.spl.2022.109682] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 11/18/2022]
|
5
|
Girard S, Stupfler G, Usseglio-Carleve A. On automatic bias reduction for extreme expectile estimation. STATISTICS AND COMPUTING 2022; 32:64. [PMID: 35968040 PMCID: PMC9362073 DOI: 10.1007/s11222-022-10118-x] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Subscribe] [Scholar Register] [Received: 01/07/2021] [Accepted: 05/30/2022] [Indexed: 06/15/2023]
Abstract
Expectiles induce a law-invariant risk measure that has recently gained popularity in actuarial and financial risk management applications. Unlike quantiles or the quantile-based Expected Shortfall, the expectile risk measure is coherent and elicitable. The estimation of extreme expectiles in the heavy-tailed framework, which is reasonable for extreme financial or actuarial risk management, is not without difficulties; currently available estimators of extreme expectiles are typically biased and hence may show poor finite-sample performance even in fairly large samples. We focus here on the construction of bias-reduced extreme expectile estimators for heavy-tailed distributions. The rationale for our construction hinges on a careful investigation of the asymptotic proportionality relationship between extreme expectiles and their quantile counterparts, as well as of the extrapolation formula motivated by the heavy-tailed context. We accurately quantify and estimate the bias incurred by the use of these relationships when constructing extreme expectile estimators. This motivates the introduction of classes of bias-reduced estimators whose asymptotic properties are rigorously shown, and whose finite-sample properties are assessed on a simulation study and three samples of real data from economics, insurance and finance. Supplementary Information The online version contains supplementary material available at 10.1007/s11222-022-10118-x.
Collapse
Affiliation(s)
- Stéphane Girard
- Univ. Grenoble Alpes, Inria, CNRS, Grenoble INP, LJK, 38000 Grenoble, France
| | - Gilles Stupfler
- Univ. Rennes, Ensai, CNRS, CREST, UMR 9194, 35000 Rennes, France
| | | |
Collapse
|
6
|
Affiliation(s)
- Dimitri Konen
- Université libre de Bruxelles, ECARES and Département de Mathématique, Avenue F.D. Roosevelt, 50, ECARES, CP114/04, B-1050, Brussels, Belgium
| | - Davy Paindaveine
- Université libre de Bruxelles, ECARES and Département de Mathématique, Avenue F.D. Roosevelt, 50, ECARES, CP114/04, B-1050, Brussels, Belgium
| |
Collapse
|
7
|
Daouia A, Gijbels I, Stupfler G. Extremile Regression. J Am Stat Assoc 2022. [DOI: 10.1080/01621459.2021.1875837] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/22/2022]
Affiliation(s)
| | - Irène Gijbels
- Statistics and Risk Section, KU Leuven, Leuven, Belgium
| | | |
Collapse
|
8
|
Padoan SA, Stupfler G. Joint inference on extreme expectiles for multivariate heavy-tailed distributions. BERNOULLI 2022. [DOI: 10.3150/21-bej1375] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/19/2022]
Affiliation(s)
- Simone A. Padoan
- Department of Decision Sciences, Bocconi University, via Roentgen 1, 20136 Milano, Italy
| | | |
Collapse
|
9
|
Stupfler G, Usseglio‐Carleve A. Composite bias‐reduced Lp‐quantile‐based estimators of extreme quantiles and expectiles. CAN J STAT 2022. [DOI: 10.1002/cjs.11703] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/07/2022]
Affiliation(s)
- Gilles Stupfler
- Department of Statistics, Univ Rennes, Ensai, CNRS CREST ‐ UMR 9194, F‐35000 Rennes France
| | - Antoine Usseglio‐Carleve
- Department of Statistics and Business Intelligence, Avignon Université LMA EA 2151, 84000 Avignon France
| |
Collapse
|
10
|
Best-Arm Identification Using Extreme Value Theory Estimates of the CVaR. JOURNAL OF RISK AND FINANCIAL MANAGEMENT 2022. [DOI: 10.3390/jrfm15040172] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 11/16/2022]
Abstract
We consider a risk-aware multi-armed bandit framework with the goal of avoiding catastrophic risk. Such a framework has multiple applications in financial risk management. We introduce a new conditional value-at-risk (CVaR) estimation procedure combining extreme value theory with automated threshold selection by ordered goodness-of-fit tests, and we apply this procedure to a pure exploration best-arm identification problem under a fixed budget. We empirically compare our results with the commonly used sample average estimator of the CVaR, and we show a significant performance improvement when the underlying arm distributions are heavy-tailed.
Collapse
|
11
|
Optimal model averaging estimator for expectile regressions. J Stat Plan Inference 2022. [DOI: 10.1016/j.jspi.2021.08.003] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/29/2022]
|
12
|
Almanjahie IM, Bouzebda S, Kaid Z, Laksaci A. Nonparametric estimation of expectile regression in functional dependent data. J Nonparametr Stat 2022. [DOI: 10.1080/10485252.2022.2027412] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/19/2022]
Affiliation(s)
- Ibrahim M. Almanjahie
- Department of Mathematics, College of Science, King Khalid University, Abha, Saudi Arabia
- Statistical Research and Studies Support Unit, King Khalid University, Abha, Saudi Arabia
| | - Salim Bouzebda
- Université de Technologie de Compiègne, L.M.A.C., Compiègne, France
| | - Zoulikha Kaid
- Department of Mathematics, College of Science, King Khalid University, Abha, Saudi Arabia
- Statistical Research and Studies Support Unit, King Khalid University, Abha, Saudi Arabia
| | - Ali Laksaci
- Department of Mathematics, College of Science, King Khalid University, Abha, Saudi Arabia
- Statistical Research and Studies Support Unit, King Khalid University, Abha, Saudi Arabia
| |
Collapse
|
13
|
Marbac M, Sedki M, Biernacki C, Vandewalle V. Simultaneous Semiparametric Estimation of Clustering and Regression. J Comput Graph Stat 2022. [DOI: 10.1080/10618600.2021.2000872] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/19/2022]
Affiliation(s)
| | | | | | - Vincent Vandewalle
- Inria, Univ. Lille, CHU Lille, ULR 2694 - METRICS: Évaluation des technologies de santé et des pratiques médicales, Lille, France
| |
Collapse
|
14
|
|
15
|
Girard S, Stupfler G, Usseglio-Carleve A. Extreme conditional expectile estimation in heavy-tailed heteroscedastic regression models. Ann Stat 2021. [DOI: 10.1214/21-aos2087] [Citation(s) in RCA: 4] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/19/2022]
|
16
|
Rachdi M, Laksaci A, Al-Kandari NM. Expectile regression for spatial functional data analysis (sFDA). METRIKA 2021. [DOI: 10.1007/s00184-021-00846-x] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/30/2022]
|
17
|
The functional kNN estimator of the conditional expectile: Uniform consistency in number of neighbors. STATISTICS & RISK MODELING 2021. [DOI: 10.1515/strm-2019-0029] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/15/2022]
Abstract
Abstract
The main purpose of the present paper is to investigate the problem of the nonparametric estimation of the expectile regression
in which the response variable is scalar while the covariate is
a random function. More precisely, an estimator is constructed by using the k Nearest Neighbor procedures (kNN). The main contribution of this study is the establishment of the Uniform consistency in Number of Neighbors (UNN) of the constructed estimator. The usefulness of our result for the smoothing parameter automatic selection is discussed. Short simulation results show that the finite sample performance of the proposed estimator is satisfactory in moderate sample sizes.
We finally examine the implementation of this model in practice with a real data in financial risk analysis.
Collapse
|
18
|
Xu Q, Ding X, Jiang C, Yu K, Shi L. An elastic-net penalized expectile regression with applications. J Appl Stat 2021; 48:2205-2230. [DOI: 10.1080/02664763.2020.1787355] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/24/2022]
Affiliation(s)
- Q.F. Xu
- School of Management, Hefei University of Technology, Hefei, People's Republic of China
- Key Laboratory of Process Optimization and Intelligent Decision-making, Ministry of Education, Hefei, People's Republic of China
| | - X.H. Ding
- School of Management, Hefei University of Technology, Hefei, People's Republic of China
| | - C.X. Jiang
- School of Management, Hefei University of Technology, Hefei, People's Republic of China
| | - K.M. Yu
- Department of Mathematics, Brunel University London, Uxbridge, UK
| | - L. Shi
- School of Computer Science and Technology, Huaibei Normal University, Huaibei, People's Republic of China
| |
Collapse
|
19
|
Linear expectile regression under massive data. FUNDAMENTAL RESEARCH 2021. [DOI: 10.1016/j.fmre.2021.08.012] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/22/2022] Open
|
20
|
Semi-parametric estimation of multivariate extreme expectiles. J MULTIVARIATE ANAL 2021. [DOI: 10.1016/j.jmva.2021.104758] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/15/2022]
|
21
|
Soale AN, Dong Y. On expectile-assisted inverse regression estimation for sufficient dimension reduction. J Stat Plan Inference 2021. [DOI: 10.1016/j.jspi.2020.11.004] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 12/17/2022]
|
22
|
Mohammedi M, Bouzebda S, Laksaci A. The consistency and asymptotic normality of the kernel type expectile regression estimator for functional data. J MULTIVARIATE ANAL 2021. [DOI: 10.1016/j.jmva.2020.104673] [Citation(s) in RCA: 4] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/23/2022]
|
23
|
Girard S, Stupfler G, Usseglio‐Carleve A. Nonparametric extreme conditional expectile estimation. Scand Stat Theory Appl 2020. [DOI: 10.1111/sjos.12502] [Citation(s) in RCA: 4] [Impact Index Per Article: 0.8] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/28/2022]
Affiliation(s)
- Stéphane Girard
- Inria, CNRS, Grenoble INP, LJK Univ. Grenoble Alpes Grenoble France
| | - Gilles Stupfler
- Ensai, CNRS, CREST ‐ UMR 9194 Univ. Rennes Rennes France
- School of Mathematical Sciences University of Nottingham Nottingham UK
| | - Antoine Usseglio‐Carleve
- Inria, CNRS, Grenoble INP, LJK Univ. Grenoble Alpes Grenoble France
- Ensai, CNRS, CREST ‐ UMR 9194 Univ. Rennes Rennes France
| |
Collapse
|
24
|
|
25
|
Gardes L, Girard S, Stupfler G. Beyond tail median and conditional tail expectation: Extreme risk estimation using tail
L
p
‐optimization. Scand Stat Theory Appl 2019. [DOI: 10.1111/sjos.12433] [Citation(s) in RCA: 4] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/26/2022]
Affiliation(s)
| | - Stéphane Girard
- Université Grenoble Alpes, Inria, CNRS, Grenoble INP, LJK 38000 Grenoble France
| | | |
Collapse
|
26
|
Pele DT, Lazar E, Mazurencu-Marinescu-Pele M. Modeling Expected Shortfall Using Tail Entropy. ENTROPY 2019; 21:1204. [PMCID: PMC7514549 DOI: 10.3390/e21121204] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.2] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Subscribe] [Scholar Register] [Received: 10/10/2019] [Accepted: 12/05/2019] [Indexed: 06/16/2023]
Abstract
Given the recent replacement of value-at-risk as the regulatory standard measure of risk with expected shortfall (ES) undertaken by the Basel Committee on Banking Supervision, it is imperative that ES gives correct estimates for the value of expected levels of losses in crisis situations. However, the measurement of ES is affected by a lack of observations in the tail of the distribution. While kernel-based smoothing techniques can be used to partially circumvent this problem, in this paper we propose a simple nonparametric tail measure of risk based on information entropy and compare its backtesting performance with that of other standard ES models.
Collapse
Affiliation(s)
- Daniel Traian Pele
- Department of Statistics and Econometrics, Faculty of Cybernetics, Statistics and Economic Informatics, The Bucharest University of Economic Studies, Piata Romana, nr.6, Sector 1, 010371 Bucharest, Romania;
| | - Emese Lazar
- Henley Business School, University of Reading, ICMA Centre, Whiteknights, Reading RG6 6BA, UK;
| | - Miruna Mazurencu-Marinescu-Pele
- Department of Statistics and Econometrics, Faculty of Cybernetics, Statistics and Economic Informatics, The Bucharest University of Economic Studies, Piata Romana, nr.6, Sector 1, 010371 Bucharest, Romania;
| |
Collapse
|
27
|
A New Parameter Estimator for the Generalized Pareto Distribution under the Peaks over Threshold Framework. MATHEMATICS 2019. [DOI: 10.3390/math7050406] [Citation(s) in RCA: 10] [Impact Index Per Article: 1.7] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 11/16/2022]
Abstract
Techniques used to analyze exceedances over a high threshold are in great demand for research in economics, environmental science, and other fields. The generalized Pareto distribution (GPD) has been widely used to fit observations exceeding the tail threshold in the peaks over threshold (POT) framework. Parameter estimation and threshold selection are two critical issues for threshold-based GPD inference. In this work, we propose a new GPD-based estimation approach by combining the method of moments and likelihood moment techniques based on the least squares concept, in which the shape and scale parameters of the GPD can be simultaneously estimated. To analyze extreme data, the proposed approach estimates the parameters by minimizing the sum of squared deviations between the theoretical GPD function and its expectation. Additionally, we introduce a recently developed stopping rule to choose the suitable threshold above which the GPD asymptotically fits the exceedances. Simulation studies show that the proposed approach performs better or similar to existing approaches, in terms of bias and the mean square error, in estimating the shape parameter. In addition, the performance of three threshold selection procedures is assessed by estimating the value-at-risk (VaR) of the GPD. Finally, we illustrate the utilization of the proposed method by analyzing air pollution data. In this analysis, we also provide a detailed guide regarding threshold selection.
Collapse
|
28
|
Tran NM, Burdejová P, Ospienko M, Härdle WK. Principal component analysis in an asymmetric norm. J MULTIVARIATE ANAL 2019. [DOI: 10.1016/j.jmva.2018.10.004] [Citation(s) in RCA: 27] [Impact Index Per Article: 4.5] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/26/2022]
|
29
|
Daouia A, Girard S, Stupfler G. Extreme M-quantiles as risk measures: From $L^{1}$ to $L^{p}$ optimization. BERNOULLI 2019. [DOI: 10.3150/17-bej987] [Citation(s) in RCA: 21] [Impact Index Per Article: 3.5] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/19/2022]
|
30
|
Zhao X, Cheng W, Zhang P. Extreme tail risk estimation with the generalized Pareto distribution under the peaks-over-threshold framework. COMMUN STAT-THEOR M 2018. [DOI: 10.1080/03610926.2018.1549253] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.1] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/27/2022]
Affiliation(s)
- Xu Zhao
- College of Applied Sciences, Beijing University of Technology, Beijing, China
| | - Weihu Cheng
- College of Applied Sciences, Beijing University of Technology, Beijing, China
| | - Pengyue Zhang
- Department of Biomedical Informatics, College of Medicine, The Ohio State University, Columbus, Ohio, USA
| |
Collapse
|
31
|
Affiliation(s)
- Abdelaati Daouia
- Toulouse School of Economics, University of Toulouse Capitole, Toulouse, France
| | - Irène Gijbels
- Department of Mathematics and Leuven Statistics Research Centre, Katholieke Universiteit Leuven, Leuven, Belgium
| | - Gilles Stupfler
- School of Mathematical Sciences, University of Nottingham, Nottingham, United Kingdom
| |
Collapse
|
32
|
Usseglio-Carleve A. Estimation of conditional extreme risk measures from heavy-tailed elliptical random vectors. Electron J Stat 2018. [DOI: 10.1214/18-ejs1499] [Citation(s) in RCA: 4] [Impact Index Per Article: 0.6] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/19/2022]
|