-
Statistical Inference on a Changing Extremal Dependence Structure
Authors:
Holger Drees
Abstract:
We analyze the extreme value dependence of independent, not necessarily identically distributed multivariate regularly varying random vectors. More specifically, we propose estimators of the spectral measure locally at some time point and of the spectral measures integrated over time. The uniform asymptotic normality of these estimators is proved under suitable nonparametric smoothness and regular…
▽ More
We analyze the extreme value dependence of independent, not necessarily identically distributed multivariate regularly varying random vectors. More specifically, we propose estimators of the spectral measure locally at some time point and of the spectral measures integrated over time. The uniform asymptotic normality of these estimators is proved under suitable nonparametric smoothness and regularity assumptions. We then use the process convergence of the integrated spectral measure to devise consistent tests for the null hypothesis that the spectral measure does not change over time. The finite sample performance of these tests is investigated in Monte Carlo simulations.
△ Less
Submitted 2 June, 2023; v1 submitted 17 January, 2022;
originally announced January 2022.
-
Cluster based inference for extremes of time series
Authors:
Holger Drees,
Anja Janßen,
Sebastian Neblung
Abstract:
We introduce a new type of estimator for the spectral tail process of a regularly varying time series. The approach is based on a characterizing invariance property of the spectral tail process, which is incorporated into the new estimator via a projection technique. We show uniform asymptotic normality of this estimator, both in the case of known and of unknown index of regular variation. In a si…
▽ More
We introduce a new type of estimator for the spectral tail process of a regularly varying time series. The approach is based on a characterizing invariance property of the spectral tail process, which is incorporated into the new estimator via a projection technique. We show uniform asymptotic normality of this estimator, both in the case of known and of unknown index of regular variation. In a simulation study the new procedure shows a more stable performance than previously proposed estimators.
△ Less
Submitted 15 March, 2021;
originally announced March 2021.
-
Dual Dynamic Programming for Multi-Scale Mixed-Integer MPC
Authors:
Ranjeet Kumar,
Michael J. Wenzel,
Mohammad N. ElBsat,
Michael J. Risbeck,
Kirk H. Drees,
Victor M. Zavala
Abstract:
We propose a dual dynamic integer programming (DDIP) framework for solving multi-scale mixed-integer model predictive control (MPC) problems. Such problems arise in applications that involve long horizons and/or fine temporal discretizations as well as mixed-integer states and controls (e.g., scheduling logic and discrete actuators). The approach uses a nested cutting-plane scheme that performs fo…
▽ More
We propose a dual dynamic integer programming (DDIP) framework for solving multi-scale mixed-integer model predictive control (MPC) problems. Such problems arise in applications that involve long horizons and/or fine temporal discretizations as well as mixed-integer states and controls (e.g., scheduling logic and discrete actuators). The approach uses a nested cutting-plane scheme that performs forward and backward sweeps along the time horizon to adaptively approximate cost-to-go functions. The DDIP scheme proposed can handle general MPC formulations with mixed-integer controls and states and can perform forward-backward sweeps over block time partitions. We demonstrate the performance of the proposed scheme by solving mixed-integer MPC problems that arise in the scheduling of central heating, ventilation, and air-conditioning (HVAC) plants. We show that the proposed scheme is scalable and dramatically outperforms state-of-the-art mixed-integer solvers.
△ Less
Submitted 16 July, 2020;
originally announced July 2020.
-
Asymptotics for sliding blocks estimators of rare events
Authors:
Holger Drees,
Sebastian Neblung
Abstract:
Drees and Rootzén (2010) have established limit theorems for a general class of empirical processes of statistics that are useful for the extreme value analysis of time series, but do not apply to statistics of sliding blocks, including so-called runs estimators. We generalize these results to empirical processes which cover both the class considered by Drees and Rootzén (2010) and processes of sl…
▽ More
Drees and Rootzén (2010) have established limit theorems for a general class of empirical processes of statistics that are useful for the extreme value analysis of time series, but do not apply to statistics of sliding blocks, including so-called runs estimators. We generalize these results to empirical processes which cover both the class considered by Drees and Rootzén (2010) and processes of sliding blocks statistics. Using this approach, one can analyze different types of statistics in a unified framework. We show that statistics based on sliding blocks are asymptotically normal with an asymptotic variance which, under rather mild conditions, is smaller than or equal to the asymptotic variance of the corresponding estimator based on disjoint blocks. Finally, the general theory is applied to three well-known estimators of the extremal index. It turns out that they all have the same limit distribution, a fact which has so far been overlooked in the literature.
△ Less
Submitted 1 September, 2020; v1 submitted 2 March, 2020;
originally announced March 2020.
-
Stochastic Model Predictive Control for Central HVAC Plants
Authors:
Ranjeet Kumar,
Michael J. Wenzel,
Mohammad N. ElBsat,
Michael J. Risbeck,
Kirk H. Drees,
Victor M. Zavala
Abstract:
We present a stochastic model predictive control (MPC) framework for central heating, ventilation, and air conditioning (HVAC) plants. The framework uses real data to forecast and quantify uncertainty of disturbances affecting the system over multiple timescales (electrical loads, heating/cooling loads, and energy prices). We conduct detailed closed-loop simulations and systematic benchmarks for t…
▽ More
We present a stochastic model predictive control (MPC) framework for central heating, ventilation, and air conditioning (HVAC) plants. The framework uses real data to forecast and quantify uncertainty of disturbances affecting the system over multiple timescales (electrical loads, heating/cooling loads, and energy prices). We conduct detailed closed-loop simulations and systematic benchmarks for the central HVAC plant of a typical university campus. Results demonstrate that deterministic MPC fails to properly capture disturbances and that this translates into economic penalties associated with peak demand charges and constraint violations in thermal storage capacity (overflow and/or depletion). Our results also demonstrate that stochastic MPC provides a more systematic approach to mitigate uncertainties and that this ultimately leads to cost savings of up to 7.5% and to mitigation of storage constraint violations. Benchmark results also indicate that these savings are close to ideal savings (9.6%) obtained under MPC with perfect information.
△ Less
Submitted 23 February, 2020;
originally announced February 2020.
-
Principal Component Analysis for Multivariate Extremes
Authors:
Holger Drees,
Anne Sabourin
Abstract:
The first order behavior of multivariate heavy-tailed random vectors above large radial thresholds is ruled by a limit measure in a regular variation framework. For a high dimensional vector, a reasonable assumption is that the support of this measure is concentrated on a lower dimensional subspace, meaning that certain linear combinations of the components are much likelier to be large than other…
▽ More
The first order behavior of multivariate heavy-tailed random vectors above large radial thresholds is ruled by a limit measure in a regular variation framework. For a high dimensional vector, a reasonable assumption is that the support of this measure is concentrated on a lower dimensional subspace, meaning that certain linear combinations of the components are much likelier to be large than others. Identifying this subspace and thus reducing the dimension will facilitate a refined statistical analysis. In this work we apply Principal Component Analysis (PCA) to a re-scaled version of radially thresholded observations. Within the statistical learning framework of empirical risk minimization, our main focus is to analyze the squared reconstruction error for the exceedances over large radial thresholds. We prove that the empirical risk converges to the true risk, uniformly over all projection subspaces. As a consequence, the best projection subspace is shown to converge in probability to the optimal one, in terms of the Hausdorff distance between their intersections with the unit sphere. In addition, if the exceedances are re-scaled to the unit ball, we obtain finite sample uniform guarantees to the reconstruction error pertaining to the estimated projection sub-space. Numerical experiments illustrate the relevance of the proposed framework for practical purposes.
△ Less
Submitted 26 June, 2019;
originally announced June 2019.
-
Peak-over-Threshold Estimators for Spectral Tail Processes: Random vs Deterministic Thresholds
Authors:
Holger Drees,
Miran Knezevic
Abstract:
The extreme value dependence of regularly varying stationary time series can be described by the spectral tail process. Drees, Segers and Warchol [Extremes 18(3): 369--402, 2015] proposed estimators of the marginal distributions of this process based on exceedances over high deterministic thresholds and analyzed their asymptotic behavior. In practice, however, versions of the estimators are applie…
▽ More
The extreme value dependence of regularly varying stationary time series can be described by the spectral tail process. Drees, Segers and Warchol [Extremes 18(3): 369--402, 2015] proposed estimators of the marginal distributions of this process based on exceedances over high deterministic thresholds and analyzed their asymptotic behavior. In practice, however, versions of the estimators are applied which use exceedances over random thresholds like intermediate order statistics. We prove that these modified estimators have the same limit distributions. This finding is corroborated in a simulation study, but the version using order statistics performs a bit better for finite samples.
△ Less
Submitted 22 July, 2019; v1 submitted 16 January, 2019;
originally announced January 2019.
-
On a minimum distance procedure for threshold selection in tail analysis
Authors:
Holger Drees,
Anja Janßen,
Sidney I. Resnick,
Tiandong Wang
Abstract:
Power-law distributions have been widely observed in different areas of scientific research. Practical estimation issues include how to select a threshold above which observations follow a power-law distribution and then how to estimate the power-law tail index. A minimum distance selection procedure (MDSP) is proposed in Clauset et al. (2009) and has been widely adopted in practice, especially in…
▽ More
Power-law distributions have been widely observed in different areas of scientific research. Practical estimation issues include how to select a threshold above which observations follow a power-law distribution and then how to estimate the power-law tail index. A minimum distance selection procedure (MDSP) is proposed in Clauset et al. (2009) and has been widely adopted in practice, especially in the analyses of social networks. However, theoretical justifications for this selection procedure remain scant. In this paper, we study the asymptotic behavior of the selected threshold and the corresponding power-law index given by the MDSP. We find that the MDSP tends to choose too high a threshold level and leads to Hill estimates with large variances and root mean squared errors for simulated data with Pareto-like tails.
△ Less
Submitted 12 February, 2020; v1 submitted 15 November, 2018;
originally announced November 2018.
-
Hierarchical MPC Schemes for Periodic Systems using Stochastic Programming
Authors:
Ranjeet Kumar,
Michael J. Wenzel,
Matthew J. Ellis,
Mohammad N. ElBsat,
Kirk H. Drees,
Victor M. Zavala
Abstract:
We show that stochastic programming (SP) provides a framework to design hierarchical model predictive control (MPC) schemes for periodic systems. This is based on the observation that, if the state policy of an infinite-horizon problem is periodic, the problem can be cast as a stochastic program (SP). This reveals that it is possible to update periodic state targets by solving a retroactive optimi…
▽ More
We show that stochastic programming (SP) provides a framework to design hierarchical model predictive control (MPC) schemes for periodic systems. This is based on the observation that, if the state policy of an infinite-horizon problem is periodic, the problem can be cast as a stochastic program (SP). This reveals that it is possible to update periodic state targets by solving a retroactive optimization problem that progressively accumulates historical data. Moreover, we show that the retroactive problem is a statistical approximation of the SP and thus delivers optimal targets in the long run. Notably, this optimality property can be achieved without the need for data forecasts and cannot be achieved by any known proactive receding horizon scheme. The SP setting also reveals that the retroactive problem can be seen as a high-level hierarchical layer that provides targets to guide a low-level MPC controller that operates over a short period at high resolution. We derive a retroactive scheme tailored to linear systems by using cutting plane techniques and suggest strategies to handle nonlinear systems.
△ Less
Submitted 28 April, 2018;
originally announced April 2018.
-
Extreme Value Estimation for Discretely Sampled Continuous Processes
Authors:
Holger Drees,
Laurens de Haan,
Feridun Turkman
Abstract:
In environmental applications of extreme value statistics, the underlying stochastic process is often modeled either as a max-stable process in continuous time/space or as a process in the domain of attraction of such a max-stable process. In practice, however, the processes are typically only observed at discrete points and one has to resort to interpolation to fill in the gaps. We discuss the in…
▽ More
In environmental applications of extreme value statistics, the underlying stochastic process is often modeled either as a max-stable process in continuous time/space or as a process in the domain of attraction of such a max-stable process. In practice, however, the processes are typically only observed at discrete points and one has to resort to interpolation to fill in the gaps. We discuss the influence of such an interpolation on estimators of marginal parameters as well as estimators of the exponent measure. In particular, natural conditions on the fineness of the observational scheme are developed which ensure that asymptotically the interpolated estimators behave in the same way as the estimators which use fully observed continuous processes.
△ Less
Submitted 9 February, 2018; v1 submitted 20 September, 2017;
originally announced September 2017.
-
Conditional Extreme Value Models: Fallacies and Pitfalls
Authors:
Holger Drees,
Anja Janßen
Abstract:
Conditional extreme value models have been introduced by Heffernan and Resnick (2007) to describe the asymptotic behavior of a random vector as one specific component becomes extreme. Obviously, this class of models is related to classical multivariate extreme value theory which describes the behavior of a random vector as its norm (and therefore at least one of its components) becomes extreme. Ho…
▽ More
Conditional extreme value models have been introduced by Heffernan and Resnick (2007) to describe the asymptotic behavior of a random vector as one specific component becomes extreme. Obviously, this class of models is related to classical multivariate extreme value theory which describes the behavior of a random vector as its norm (and therefore at least one of its components) becomes extreme. However, it turns out that this relationship is rather subtle and sometimes contrary to intuition. We clarify the differences between the two approaches with the help of several illuminative (counter)examples. Furthermore, we discuss marginal standardization, which is a useful tool in classical multivariate extreme value theory but, as we point out, much less straightforward and sometimes even obscuring in conditional extreme value models. Finally, we indicate how, in some situations, a more comprehensive characterization of the asymptotic behavior can be obtained if the conditions of conditional extreme value models are relaxed so that the limit is no longer unique.
△ Less
Submitted 23 February, 2017; v1 submitted 9 June, 2016;
originally announced June 2016.
-
Bootstrap** Empirical Processes of Cluster Functionals with Application to Extremograms
Authors:
Holger Drees
Abstract:
In the extreme value analysis of time series, not only the tail behavior is of interest, but also the serial dependence plays a crucial role. Drees and Rootzén (2010) established limit theorems for a general class of empirical processes of so-called cluster functionals which can be used to analyse various aspects of the extreme value behavior of mixing time series. However, usually the limit distr…
▽ More
In the extreme value analysis of time series, not only the tail behavior is of interest, but also the serial dependence plays a crucial role. Drees and Rootzén (2010) established limit theorems for a general class of empirical processes of so-called cluster functionals which can be used to analyse various aspects of the extreme value behavior of mixing time series. However, usually the limit distribution is too complex to enable a direct construction of confidence regions. Therefore, we suggest a multiplier block bootstrap analog to the empirical processes of cluster functionals. It is shown that under virtually the same conditions as used by Drees and Rootzén (2010), conditionally on the data, the bootstrap processes converge to the same limit distribution. These general results are applied to construct confidence regions for the empirical extremogram introduced by Davis and Mikosch (2009). In a simulation study, the confidence intervals constructed by our multiplier block bootstrap approach compare favorably to the stationary bootstrap proposed by Davis et al.\ (2012).
△ Less
Submitted 2 November, 2015;
originally announced November 2015.
-
Correction note to "Limit Theorems for Empirical Processes of Cluster Functionals" [arXiv:0910.0343]
Authors:
Holger Drees,
Holger Rootzén
Abstract:
We correct an error in a technical lemma of Drees and Rootzén (2010) [arXiv:0910.0343] and discuss consequences for applications.
We correct an error in a technical lemma of Drees and Rootzén (2010) [arXiv:0910.0343] and discuss consequences for applications.
△ Less
Submitted 29 October, 2015;
originally announced October 2015.
-
Joint exceedances of random products
Authors:
Anja Janßen,
Holger Drees
Abstract:
We analyze the joint extremal behavior of $n$ random products of the form $\prod_{j=1}^m X_j^{a_{ij}}, 1 \leq i \leq n,$ for non-negative, independent regularly varying random variables $X_1, \ldots, X_m$ and general coefficients $a_{ij} \in \mathbb{R}$. Products of this form appear for example if one observes a linear time series with gamma type innovations at $n$ points in time. We combine argum…
▽ More
We analyze the joint extremal behavior of $n$ random products of the form $\prod_{j=1}^m X_j^{a_{ij}}, 1 \leq i \leq n,$ for non-negative, independent regularly varying random variables $X_1, \ldots, X_m$ and general coefficients $a_{ij} \in \mathbb{R}$. Products of this form appear for example if one observes a linear time series with gamma type innovations at $n$ points in time. We combine arguments of linear optimization and a generalized concept of regular variation on cones to show that the asymptotic behavior of joint exceedance probabilities of these products is determined by the solution of a linear program related to the matrix $\mathbf{A}=(a_{ij})$.
△ Less
Submitted 12 May, 2016; v1 submitted 13 May, 2015;
originally announced May 2015.
-
Hypotheses tests in boundary regression models
Authors:
Holger Drees,
Natalie Neumeyer,
Leonie Selk
Abstract:
Consider a nonparametric regression model with one-sided errors and regression function in a general Hölder class. We estimate the regression function via minimization of the local integral of a polynomial approximation. We show uniform rates of convergence for the simple regression estimator as well as for a smooth version. These rates carry over to mean regression models with a symmetric and bou…
▽ More
Consider a nonparametric regression model with one-sided errors and regression function in a general Hölder class. We estimate the regression function via minimization of the local integral of a polynomial approximation. We show uniform rates of convergence for the simple regression estimator as well as for a smooth version. These rates carry over to mean regression models with a symmetric and bounded error distribution. In such a setting, one obtains faster rates for irregular error distributions concentrating sufficient mass near the endpoints than for the usual regular distributions. The results are applied to prove asymptotic $\sqrt{n}$-equivalence of a residual-based (sequential) empirical distribution function to the (sequential) empirical distribution function of unobserved errors in the case of irregular error distributions. This result is remarkably different from corresponding results in mean regression with regular errors. It can readily be applied to develop goodness-of-fit tests for the error distribution. We present some examples and investigate the small sample performance in a simulation study. We further discuss asymptotically distribution-free hypotheses tests for independence of the error distribution from the points of measurement and for monotonicity of the boundary function as well.
△ Less
Submitted 11 October, 2016; v1 submitted 18 August, 2014;
originally announced August 2014.
-
A stochastic volatility model with flexible extremal dependence structure
Authors:
Anja Janssen,
Holger Drees
Abstract:
Stochastic volatility processes with heavy-tailed innovations are a well-known model for financial time series. In these models, the extremes of the log returns are mainly driven by the extremes of the i.i.d. innovation sequence which leads to a very strong form of asymptotic independence, that is, the coefficient of tail dependence is equal to $1/2$ for all positive lags. We propose an alternativ…
▽ More
Stochastic volatility processes with heavy-tailed innovations are a well-known model for financial time series. In these models, the extremes of the log returns are mainly driven by the extremes of the i.i.d. innovation sequence which leads to a very strong form of asymptotic independence, that is, the coefficient of tail dependence is equal to $1/2$ for all positive lags. We propose an alternative class of stochastic volatility models with heavy-tailed volatilities and examine their extreme value behavior. In particular, it is shown that, while lagged extreme observations are typically asymptotically independent, their coefficient of tail dependence can take on any value between $1/2$ (corresponding to exact independence) and 1 (related to asymptotic dependence). Hence, this class allows for a much more flexible extremal dependence between consecutive observations than classical SV models and can thus describe the observed clustering of financial returns more realistically. The extremal dependence structure of lagged observations is analyzed in the framework of regular variation on the cone $(0,\infty)^d$. As two auxiliary results which are of interest on their own we derive a new Breiman-type theorem about regular variation on $(0,\infty)^d$ for products of a random matrix and a regularly varying random vector and a statement about the joint extremal behavior of products of i.i.d. regularly varying random variables.
△ Less
Submitted 24 March, 2016; v1 submitted 17 October, 2013;
originally announced October 2013.
-
Bias correction for estimators of the extremal index
Authors:
Holger Drees
Abstract:
We investigate the joint asymptotic behavior of so-called blocks estimator of the extremal index, that determines the mean length of clusters of extremes, based on the exceedances over different thresholds. Due to the large bias of these estimators, the resulting estimates are usually very sensitive to the choice of the threshold and thus difficult to interpret. We propose and examine a bias corre…
▽ More
We investigate the joint asymptotic behavior of so-called blocks estimator of the extremal index, that determines the mean length of clusters of extremes, based on the exceedances over different thresholds. Due to the large bias of these estimators, the resulting estimates are usually very sensitive to the choice of the threshold and thus difficult to interpret. We propose and examine a bias correction that asymptotically removes the leading bias term while the rate of convergence of the random error is preserved.
△ Less
Submitted 5 July, 2011;
originally announced July 2011.
-
Estimating failure probabilities
Authors:
Holger Drees,
Laurens de Haan
Abstract:
In risk management, often the probability must be estimated that a random vector falls into an extreme failure set. In the framework of bivariate extreme value theory, we construct an estimator for such failure probabilities and analyze its asymptotic properties under natural conditions. It turns out that the estimation error is mainly determined by the accuracy of the statistical analysis of the…
▽ More
In risk management, often the probability must be estimated that a random vector falls into an extreme failure set. In the framework of bivariate extreme value theory, we construct an estimator for such failure probabilities and analyze its asymptotic properties under natural conditions. It turns out that the estimation error is mainly determined by the accuracy of the statistical analysis of the marginal distributions if the extreme value approximation to the dependence structure is at least as accurate as the generalized Pareto approximation to the marginal distributions. Moreover, we establish confidence intervals and briefly discuss generalizations to higher dimensions and issues arising in practical applications as well.
△ Less
Submitted 3 June, 2015; v1 submitted 4 July, 2011;
originally announced July 2011.
-
Limit theorems for empirical processes of cluster functionals
Authors:
Holger Drees,
Holger Rootzén
Abstract:
Let $(X_{n,i})_{1\le i\le n,n\in\mathbb{N}}$ be a triangular array of row-wise stationary $\mathbb{R}^d$-valued random variables. We use a "blocks method" to define clusters of extreme values: the rows of $(X_{n,i})$ are divided into $m_n$ blocks $(Y_{n,j})$, and if a block contains at least one extreme value, the block is considered to contain a cluster. The cluster starts at the first extreme va…
▽ More
Let $(X_{n,i})_{1\le i\le n,n\in\mathbb{N}}$ be a triangular array of row-wise stationary $\mathbb{R}^d$-valued random variables. We use a "blocks method" to define clusters of extreme values: the rows of $(X_{n,i})$ are divided into $m_n$ blocks $(Y_{n,j})$, and if a block contains at least one extreme value, the block is considered to contain a cluster. The cluster starts at the first extreme value in the block and ends at the last one. The main results are uniform central limit theorems for empirical processes $Z_n(f):=\frac{1}{\sqrt {nv_n}}\sum_{j=1}^{m_n}(f(Y_{n,j})-Ef(Y_{n,j})),$ for $v_n=P\{X_{n,i}\neq0\}$ and $f$ belonging to classes of cluster functionals, that is, functions of the blocks $Y_{n,j}$ which only depend on the cluster values and which are equal to 0 if $Y_{n,j}$ does not contain a cluster. Conditions for finite-dimensional convergence include $β$-mixing, suitable Lindeberg conditions and convergence of covariances. To obtain full uniform convergence, we use either "bracketing entropy" or bounds on covering numbers with respect to a random semi-metric. The latter makes it possible to bring the powerful Vapnik--Červonenkis theory to bear. Applications include multivariate tail empirical processes and empirical processes of cluster values and of order statistics in clusters. Although our main field of applications is the analysis of extreme values, the theory can be applied more generally to rare events occurring, for example, in nonparametric curve estimation.
△ Less
Submitted 18 May, 2020; v1 submitted 2 October, 2009;
originally announced October 2009.
-
Some aspects of extreme value theory under serial dependence
Authors:
Holger Drees
Abstract:
On the occasion of Laurens de Haan's 70th birthday, we discuss two aspects of the statistical inference on the extreme value behavior of time series with a particular emphasis on his important contributions. First, the performance of a direct marginal tail analysis is compared with that of a model-based approach using an analysis of residuals. Second, the importance of the extremal index as a me…
▽ More
On the occasion of Laurens de Haan's 70th birthday, we discuss two aspects of the statistical inference on the extreme value behavior of time series with a particular emphasis on his important contributions. First, the performance of a direct marginal tail analysis is compared with that of a model-based approach using an analysis of residuals. Second, the importance of the extremal index as a measure of the serial extremal dependence is discussed by the example of solutions of a stochastic recurrence equation.
△ Less
Submitted 31 October, 2007;
originally announced October 2007.
-
On maximum likelihood estimation of the extreme value index
Authors:
Holger Drees,
Ana Ferreira,
Laurens de Haan
Abstract:
We prove asymptotic normality of the so-called maximum likelihood estimator of the extreme value index.
We prove asymptotic normality of the so-called maximum likelihood estimator of the extreme value index.
△ Less
Submitted 5 July, 2004;
originally announced July 2004.
