-
Distributional stability of sparse inverse covariance matrix estimators
Authors:
Renjie Chen,
Huifu Xu,
Henryk Zähle
Abstract:
Finding an approximation of the inverse of the covariance matrix, also known as precision matrix, of a random vector with empirical data is widely discussed in finance and engineering. In data-driven problems, empirical data may be ``contaminated''. This raises the question as to whether the approximate precision matrix is reliable from a statistical point of view. In this paper, we concentrate on…
▽ More
Finding an approximation of the inverse of the covariance matrix, also known as precision matrix, of a random vector with empirical data is widely discussed in finance and engineering. In data-driven problems, empirical data may be ``contaminated''. This raises the question as to whether the approximate precision matrix is reliable from a statistical point of view. In this paper, we concentrate on a much-noticed sparse estimator of the precision matrix and investigate the issue from the perspective of distributional stability. Specifically, we derive an explicit local Lipschitz bound for the distance between the distributions of the sparse estimator under two different distributions (regarded as the true data distribution and the distribution of ``contaminated'' data). The distance is measured by the Kantorovich metric on the set of all probability measures on a matrix space. We also present analogous results for the standard estimators of the covariance matrix and its eigenvalues. Furthermore, we discuss two applications and conduct some numerical experiments.
△ Less
Submitted 6 July, 2024;
originally announced July 2024.
-
First-order sensitivity of the optimal value in a Markov decision model with respect to deviations in the transition probability function
Authors:
Patrick Kern,
Axel Simroth,
Henryk Zähle
Abstract:
Markov decision models (MDM) used in practical applications are most often less complex than the underlying `true' MDM. The reduction of model complexity is performed for several reasons. However, it is obviously of interest to know what kind of model reduction is reasonable (in regard to the optimal value) and what kind is not. In this article we propose a way how to address this question. We int…
▽ More
Markov decision models (MDM) used in practical applications are most often less complex than the underlying `true' MDM. The reduction of model complexity is performed for several reasons. However, it is obviously of interest to know what kind of model reduction is reasonable (in regard to the optimal value) and what kind is not. In this article we propose a way how to address this question. We introduce a sort of derivative of the optimal value as a function of the transition probabilities, which can be used to measure the (first-order) sensitivity of the optimal value w.r.t.\ changes in the transition probabilities. `Differentiability' is obtained for a fairly broad class of MDMs, and the `derivative' is specified explicitly. Our theoretical findings are illustrated by means of optimization problems in inventory control and mathematical finance.
△ Less
Submitted 17 September, 2019;
originally announced September 2019.
-
Functional delta-method for the bootstrap of uniformly quasi-Hadamard differentiable functionals
Authors:
Eric Beutner,
Henryk Zähle
Abstract:
The functional delta-method provides a convenient tool for deriving bootstrap consistency of a sequence of plug-in estimators w.r.t. a given functional from bootstrap consistency of the underlying sequence of estimators. It has recently been shown in Beutner and Zähle (2016) that the range of applications of the functional delta-method for establishing bootstrap consistency in probability of the s…
▽ More
The functional delta-method provides a convenient tool for deriving bootstrap consistency of a sequence of plug-in estimators w.r.t. a given functional from bootstrap consistency of the underlying sequence of estimators. It has recently been shown in Beutner and Zähle (2016) that the range of applications of the functional delta-method for establishing bootstrap consistency in probability of the sequence of plug-in estimators can be considerably enlarged by replacing the usual condition of Hadamard differentiability of the given functional by the weaker condition of quasi-Hadamard differentiability. Here we introduce the notion of uniform quasi-Hadamard differentiability and show that this notion extends the set of functionals for which almost sure bootstrap consistency of the corresponding sequence of plug-in estimators can be obtained by the functional delta-method. We illustrate the benefit of our results by means of the Average Value at Risk functional as well as the composition of the Average Value at Risk functional and the compound convolution functional. For the latter we use a chain rule to be proved here. In our examples we consider the weighted exchangeable bootstrap for independent observations and the blockwise bootstrap for $β$-mixing observations.
△ Less
Submitted 20 September, 2016; v1 submitted 19 September, 2016;
originally announced September 2016.
-
Statistical inference for expectile-based risk measures
Authors:
Volker Krätschmer,
Henryk Zähle
Abstract:
Expectiles were introduced by Newey and Powell (1987) in the context of linear regression models. Recently, Bellini et al. (2014) revealed that expectiles can also be seen as reasonable law-invariant risk measures. In this article, we show that the corresponding statistical functionals are continuous w.r.t. the $1$-weak topology and suitably functionally differentiable. By means of these regularit…
▽ More
Expectiles were introduced by Newey and Powell (1987) in the context of linear regression models. Recently, Bellini et al. (2014) revealed that expectiles can also be seen as reasonable law-invariant risk measures. In this article, we show that the corresponding statistical functionals are continuous w.r.t. the $1$-weak topology and suitably functionally differentiable. By means of these regularity results we can derive several properties as consistency, asymptotic normality, bootstrap consistency, and qualitative robustness of the corresponding estimators in nonparametric and parametric statistical models.
△ Less
Submitted 20 September, 2016; v1 submitted 20 January, 2016;
originally announced January 2016.
-
Domains of weak continuity of statistical functionals with a view toward robust statistics
Authors:
Volker Krätschmer,
Alexander Schied,
Henryk Zähle
Abstract:
Many standard estimators such as several maximum likelihood estimators or the empirical estimator for any law-invariant convex risk measure are not (qualitatively) robust in the classical sense. However, these estimators may nevertheless satisfy a weak (Krätschmer/Schied/Zähle 2012, 2014) or a local (Zähle 2016) robustness property on relevant sets of distributions. One aim of our paper is to iden…
▽ More
Many standard estimators such as several maximum likelihood estimators or the empirical estimator for any law-invariant convex risk measure are not (qualitatively) robust in the classical sense. However, these estimators may nevertheless satisfy a weak (Krätschmer/Schied/Zähle 2012, 2014) or a local (Zähle 2016) robustness property on relevant sets of distributions. One aim of our paper is to identify sets of local robustness, and to explain the benefit of the knowledge of such sets. For instance, we will be able to demonstrate that many maximum likelihood estimators are robust on their natural parametric domains. A second aim consists in extending the general theory of robust estimation to our local framework. In particular we provide a corresponding Hampel-type theorem linking local robustness of a plug-in estimator with a certain continuity condition.
△ Less
Submitted 20 June, 2016; v1 submitted 27 November, 2015;
originally announced November 2015.
-
Functional delta-method for the bootstrap of quasi-Hadamard differentiable functionals
Authors:
Eric Beutner,
Henryk Zähle
Abstract:
The functional delta-method provides a convenient tool for deriving the asymptotic distribution of a plug-in estimator of a statistical functional from the asymptotic distribution of the respective empirical process. Moreover, it provides a tool to derive bootstrap consistency for plug-in estimators from bootstrap consistency of empirical processes. It has recently been shown that the range of app…
▽ More
The functional delta-method provides a convenient tool for deriving the asymptotic distribution of a plug-in estimator of a statistical functional from the asymptotic distribution of the respective empirical process. Moreover, it provides a tool to derive bootstrap consistency for plug-in estimators from bootstrap consistency of empirical processes. It has recently been shown that the range of applications of the functional delta-method for the asymptotic distribution can be considerably enlarged by employing the notion of quasi-Hadamard differentiability. Here we show in a general setting that this enlargement carries over to the bootstrap. That is, for quasi-Hadamard differentiable functionals bootstrap consistency of the plug-in estimator follows from bootstrap consistency of the respective empirical process. This enlargement often requires convergence in distribution of the bootstrapped empirical process w.r.t.\ a nonuniform sup-norm. The latter is not problematic as will be illustrated by means of examples.
△ Less
Submitted 4 May, 2016; v1 submitted 21 October, 2015;
originally announced October 2015.
-
Nonparametric estimation of risk measures of collective risks
Authors:
Alexandra Lauer,
Henryk Zähle
Abstract:
We consider two nonparametric estimators for the risk measure of the sum of $n$ i.i.d. individual insurance risks where the number of historical single claims that are used for the statistical estimation is of order $n$. This framework matches the situation that nonlife insurance companies are faced with within in the scope of premium calculation. Indeed, the risk measure of the aggregate risk div…
▽ More
We consider two nonparametric estimators for the risk measure of the sum of $n$ i.i.d. individual insurance risks where the number of historical single claims that are used for the statistical estimation is of order $n$. This framework matches the situation that nonlife insurance companies are faced with within in the scope of premium calculation. Indeed, the risk measure of the aggregate risk divided by $n$ can be seen as a suitable premium for each of the individual risks. For both estimators divided by $n$ we derive a sort of Marcinkiewicz--Zygmund strong law as well as a weak limit theorem. The behavior of the estimators for small to moderate $n$ is studied by means of Monte-Carlo simulations.
△ Less
Submitted 16 September, 2015; v1 submitted 10 April, 2015;
originally announced April 2015.
-
On qualitative robustness of the Lotka--Nagaev estimator for the offspring mean of a supercritical Galton--Watson process
Authors:
Dominic Schuhmacher,
Anja Sturm,
Henryk Zähle
Abstract:
We characterize the sets of offspring laws on which the Lotka--Nagaev estimator for the mean of a supercritical Galton--Watson process is qualitatively robust. These are exactly the locally uniformly integrating sets of offspring laws, which may be quite large. If the corresponding global property is assumed instead, we obtain uniform robustness as well. We illustrate both results with a number…
▽ More
We characterize the sets of offspring laws on which the Lotka--Nagaev estimator for the mean of a supercritical Galton--Watson process is qualitatively robust. These are exactly the locally uniformly integrating sets of offspring laws, which may be quite large. If the corresponding global property is assumed instead, we obtain uniform robustness as well. We illustrate both results with a number of concrete examples. As a by-product of the proof we obtain that the Lotka--Nagaev estimator is [locally] uniformly weakly consistent on the respective sets of offspring laws, conditionally on non-extinction.
△ Less
Submitted 13 February, 2015; v1 submitted 15 September, 2014;
originally announced September 2014.
-
A definition of qualitative robustness for general point estimators, and examples
Authors:
Henryk Zähle
Abstract:
A definition of qualitative robustness for point estimators in general statistical models is proposed. Some criteria for robustness are established and applied to estimators in parametric, semiparametric, and nonparametric models. In specific nonparametric models, the proposed definition boils down to Hampel robustness. It is also explained how plug-in estimators in certain nonparametric models ca…
▽ More
A definition of qualitative robustness for point estimators in general statistical models is proposed. Some criteria for robustness are established and applied to estimators in parametric, semiparametric, and nonparametric models. In specific nonparametric models, the proposed definition boils down to Hampel robustness. It is also explained how plug-in estimators in certain nonparametric models can be reasonably classified w.r.t. their degrees of robustness.
△ Less
Submitted 17 June, 2015; v1 submitted 30 June, 2014;
originally announced June 2014.
-
Quasi-Hadamard differentiability of general risk functionals and its application
Authors:
Volker Krätschmer,
Alexander Schied,
Henryk Zähle
Abstract:
We apply a suitable modification of the functional delta method to statistical functionals that arise from law-invariant coherent risk measures. To this end we establish differentiability of the statistical functional in a relaxed Hadamard sense, namely with respect to a suitably chosen norm and in the directions of a specifically chosen "tangent space". We show that this notion of quasi-Hadamard…
▽ More
We apply a suitable modification of the functional delta method to statistical functionals that arise from law-invariant coherent risk measures. To this end we establish differentiability of the statistical functional in a relaxed Hadamard sense, namely with respect to a suitably chosen norm and in the directions of a specifically chosen "tangent space". We show that this notion of quasi-Hadamard differentiability yields both strong laws and limit theorems for the asymptotic distribution of the plug-in estimators. Our results can be regarded as a contribution to the statistics and numerics of risk measurement and as a case study for possible refinements of the functional delta method through fine-tuning the underlying notion of differentiability
△ Less
Submitted 16 February, 2015; v1 submitted 14 January, 2014;
originally announced January 2014.
-
Marcinkiewicz-Zygmund and ordinary strong laws for empirical distribution functions and plug-in estimators
Authors:
Henryk Zähle
Abstract:
Both Marcinkiewicz-Zygmund strong laws of large numbers (MZ-SLLNs) and ordinary strong laws of large numbers (SLLNs) for plug-in estimators of general statistical functionals are derived. It is used that if a statistical functional is "sufficiently regular", then a (MZ-) SLLN for the estimator of the unknown distribution function yields a (MZ-) SLLN for the corresponding plug-in estimator. It is i…
▽ More
Both Marcinkiewicz-Zygmund strong laws of large numbers (MZ-SLLNs) and ordinary strong laws of large numbers (SLLNs) for plug-in estimators of general statistical functionals are derived. It is used that if a statistical functional is "sufficiently regular", then a (MZ-) SLLN for the estimator of the unknown distribution function yields a (MZ-) SLLN for the corresponding plug-in estimator. It is in particular shown that many L-, V- and risk functionals are "sufficiently regular", and that known results on the strong convergence of the empirical process of α-mixing random variables can be improved. The presented approach does not only cover some known results but also provides some new strong laws for plug-in estimators of particular statistical functionals.
△ Less
Submitted 4 January, 2013;
originally announced January 2013.
-
Deriving the asymptotic distribution of U- and V-statistics of dependent data using weighted empirical processes
Authors:
Eric Beutner,
Henryk Zähle
Abstract:
It is commonly acknowledged that V-functionals with an unbounded kernel are not Hadamard differentiable and that therefore the asymptotic distribution of U- and V-statistics with an unbounded kernel cannot be derived by the Functional Delta Method (FDM). However, in this article we show that V-functionals are quasi-Hadamard differentiable and that therefore a modified version of the FDM (introduce…
▽ More
It is commonly acknowledged that V-functionals with an unbounded kernel are not Hadamard differentiable and that therefore the asymptotic distribution of U- and V-statistics with an unbounded kernel cannot be derived by the Functional Delta Method (FDM). However, in this article we show that V-functionals are quasi-Hadamard differentiable and that therefore a modified version of the FDM (introduced recently in (J. Multivariate Anal. 101 (2010) 2452--2463)) can be applied to this problem. The modified FDM requires weak convergence of a weighted version of the underlying empirical process. The latter is not problematic since there exist several results on weighted empirical processes in the literature; see, for example, (J. Econometrics 130 (2006) 307--335, Ann. Probab. 24 (1996) 2098--2127, Empirical Processes with Applications to Statistics (1986) Wiley, Statist. Sinica 18 (2008) 313--333). The modified FDM approach has the advantage that it is very flexible w.r.t. both the underlying data and the estimator of the unknown distribution function. Both will be demonstrated by various examples. In particular, we will show that our FDM approach covers mainly all the results known in literature for the asymptotic distribution of U- and V-statistics based on dependent data -- and our assumptions are by tendency even weaker. Moreover, using our FDM approach we extend these results to dependence concepts that are not covered by the existing literature.
△ Less
Submitted 25 July, 2012;
originally announced July 2012.
-
Comparative and qualitative robustness for law-invariant risk measures
Authors:
Volker Krätschmer,
Alexander Schied,
Henryk Zähle
Abstract:
When estimating the risk of a P&L from historical data or Monte Carlo simulation, the robustness of the estimate is important. We argue here that Hampel's classical notion of qualitative robustness is not suitable for risk measurement and we propose and analyze a refined notion of robustness that applies to tail-dependent law-invariant convex risk measures on Orlicz space. This concept of robustne…
▽ More
When estimating the risk of a P&L from historical data or Monte Carlo simulation, the robustness of the estimate is important. We argue here that Hampel's classical notion of qualitative robustness is not suitable for risk measurement and we propose and analyze a refined notion of robustness that applies to tail-dependent law-invariant convex risk measures on Orlicz space. This concept of robustness captures the tradeoff between robustness and sensitivity and can be quantified by an index of qualitative robustness. By means of this index, we can compare various risk measures, such as distortion risk measures, in regard to their degree of robustness. Our analysis also yields results that are of independent interest such as continuity properties and consistency of estimators for risk measures, or a Skorohod representation theorem for ψ-weak convergence.
△ Less
Submitted 14 January, 2014; v1 submitted 11 April, 2012;
originally announced April 2012.
-
Qualitative robustness of statistical functionals under strong mixing
Authors:
Henryk Zähle
Abstract:
A new concept of (asymptotic) qualitative robustness for plug-in estimators based on identically distributed possibly dependent observations is introduced, and it is shown that Hampel's theorem for general metrics $d$ still holds. Since Hampel's theorem assumes the UGC property w.r.t. $d$, that is, convergence in probability of the empirical probability measure to the true marginal distribution w.…
▽ More
A new concept of (asymptotic) qualitative robustness for plug-in estimators based on identically distributed possibly dependent observations is introduced, and it is shown that Hampel's theorem for general metrics $d$ still holds. Since Hampel's theorem assumes the UGC property w.r.t. $d$, that is, convergence in probability of the empirical probability measure to the true marginal distribution w.r.t. $d$ uniformly in the class of all admissible laws on the sample path space, this property is shown for a large class of strongly mixing laws for three different metrics $d$. For real-valued observations, the UGC property is established for both the Kolomogorov $φ$-metric and the Lévy $ψ$-metric, and for observations in a general locally compact and second countable Hausdorff space the UGC property is established for a certain metric generating the $ψ$-weak topology. The key is a new uniform weak LLN for strongly mixing random variables. The latter is of independent interest and relies on Rio's maximal inequality.
△ Less
Submitted 27 July, 2015; v1 submitted 23 March, 2012;
originally announced March 2012.
-
Continuous mapping approach to the asymptotics of $U$- and $V$-statistics
Authors:
Eric Beutner,
Henryk Zähle
Abstract:
We derive a new representation for $U$- and $V$-statistics. Using this representation, the asymptotic distribution of $U$- and $V$-statistics can be derived by a direct application of the Continuous Mapping theorem. That novel approach not only encompasses most of the results on the asymptotic distribution known in literature, but also allows for the first time a unifying treatment of non-degenera…
▽ More
We derive a new representation for $U$- and $V$-statistics. Using this representation, the asymptotic distribution of $U$- and $V$-statistics can be derived by a direct application of the Continuous Mapping theorem. That novel approach not only encompasses most of the results on the asymptotic distribution known in literature, but also allows for the first time a unifying treatment of non-degenerate and degenerate $U$- and $V$-statistics. Moreover, it yields a new and powerful tool to derive the asymptotic distribution of very general $U$- and $V$-statistics based on long-memory sequences. This will be exemplified by several astonishing examples. In particular, we shall present examples where weak convergence of $U$- or $V$-statistics occurs at the rate $a_n^3$ and $a_n^4$, respectively, when $a_n$ is the rate of weak convergence of the empirical process. We also introduce the notion of asymptotic (non-) degeneracy which often appears in the presence of long-memory sequences.
△ Less
Submitted 12 March, 2014; v1 submitted 6 March, 2012;
originally announced March 2012.
