arXiv subjects

Claudia Czado

Publications and source records attributed to Claudia Czado.

At least 37 records · Page 2Linked to original sources

arXiv2019

A Bayesian Non-linear State Space Copula Model to Predict Air Pollution in Beijing

Air pollution is a serious issue that currently affects many industrial cities in the world and can cause severe illness to the population. In particular, it has been proven that extreme high levels of airborne contaminants have dangerous short-term effects on human health, in terms of increased hospital admissions for cardiovascular and respiratory diseases and increased mortality risk. For these reasons, accurate estimation and prediction of airborne pollutant concentration is crucial. In this paper, we propose a flexible novel approach to model hourly measurements of fine particulate matter and meteorological data collected in Beijing in 2014. We show that the standard state space model, based on Gaussian assumptions, does not correctly capture the time dynamics of the observations. Therefore, we propose a non-linear non-Gaussian state space model where both the observation and the state equations are defined by copula specifications, and we perform Bayesian inference using the Hamiltonian Monte Carlo method. The proposed copula state space approach is very flexible, since it allows us to separately model the marginals and to accommodate a wide variety of dependence structures in the data dynamics. We show that the proposed approach allows us not only to predict particulate matter measurements, but also to investigate the effects of user specified climate scenarios.

stat.AP↗

arXiv2019

Bayesian inference for dynamic vine copulas in higher dimensions

We propose a class of dynamic vine copula models. This is an extension of static vine copulas and a generalization of dynamic C-vine and D-vine copulas studied by Almeida et al (2016) and Goel and Mehra (2019). Within this class, we allow for time-varying dependence by driving the vine copula parameters with latent AR(1) processes. This modeling approach is very flexible but estimation is not straightforward due to the high-dimensional parameter space. We propose a Bayesian estimation approach, which relies on a novel approximation of the posterior distribution. This approximation allows to use Markov Chain Monte Carlo methods, such as elliptical slice sampling, in a sequential way. In contrast to other Bayesian sequential estimation procedures for vine copula models as proposed by Gruber and Czado (2015), there is no need to collapse copula parameters to point estimates before proceeding to the next tree. Thus more information and uncertainty is propagated from lower to higher trees. A simulation study shows satisfactory performance of the Bayesian procedure. This dynamic modeling and inference approach can be applied in various fields, where static vine copulas have already proven to be successful, including environmental sciences, medicine and finance. Here we study the dependence among 21 exchange rates. For comparison we also estimate a static vine copula model and dynamic C-vine and D-vine copula models. This comparison shows superior performance of the proposed dynamic vine copula model with respect to one day ahead forecasting accuracy.

stat.ME↗

arXiv2019

Bayesian Multivariate Nonlinear State Space Copula Models

In this paper we propose a flexible class of multivariate nonlinear non-Gaussian state space models, based on copulas. More precisely, we assume that the observation equation and the state equation are defined by copula families that are not necessarily equal. For each time point, the resulting model can be described by a C-vine copula truncated after the first tree, where the root node is represented by the latent state. Inference is performed within the Bayesian framework, using the Hamiltonian Monte Carlo method, where a further D-vine truncated after the first tree is used as prior distribution to capture the temporal dependence in the latent states. Simulation studies show that the proposed copula-based approach is extremely flexible, since it is able to describe a wide range of dependence structures and, at the same time, allows us to deal with missing data. The application to atmospheric pollutant measurement data shows that our approach is suitable for accurate modeling and prediction of data dynamics in the presence of missing values. Comparison to a Gaussian linear state space model and to Bayesian additive regression trees shows the superior performance of the proposed model with respect to predictive accuracy.

stat.ME↗

arXiv2019

Efficient Bayesian inference for nonlinear state space models with univariate autoregressive state equation

Latent autoregressive processes are a popular choice to model time varying parameters. These models can be formulated as nonlinear state space models for which inference is not straightforward due to the high number of parameters. Therefore maximum likelihood methods are often infeasible and researchers rely on alternative techniques, such as Gibbs sampling. But conventional Gibbs samplers are often tailored to specific situations and suffer from high autocorrelation among repeated draws. We present a Gibbs sampler for general nonlinear state space models with an univariate autoregressive state equation. For this we employ an interweaving strategy and elliptical slice sampling to exploit the dependence implied by the autoregressive process. Within a simulation study we demonstrate the efficiency of the proposed sampler for bivariate dynamic copula models. Further we are interested in modeling the volatility return relationship. Therefore we use the proposed sampler to estimate the parameters of stochastic volatility models with skew Student t errors and the parameters of a novel bivariate dynamic mixture copula model. This model allows for dynamic asymmetric tail dependence. Comparison to relevant benchmark models, such as the DCC-GARCH or a Student t copula model, with respect to predictive accuracy shows the superior performance of the proposed approach.

stat.CO↗

arXiv2019

Bayesian inference for a single factor copula stochastic volatility model using Hamiltonian Monte Carlo

For modeling multivariate financial time series we propose a single factor copula model together with stochastic volatility margins. This model generalizes single factor models relying on the multivariate normal distribution and allows for symmetric and asymmetric tail dependence. We develop joint Bayesian inference using Hamiltonian Monte Carlo (HMC) within Gibbs sampling. Thus we avoid information loss caused by the two-step approach for margins and dependence in copula models as followed by Schamberger et al(2017). Further, the Bayesian approach allows for high dimensional parameter spaces as they are present here in addition to uncertainty quantification through credible intervals. By allowing for indicators for different copula families the copula families are selected automatically in the Bayesian framework. In a first simulation study the performance of HMC is compared to the Markov Chain Monte Carlo (MCMC) approach developed by Schamberger et al(2017) for the copula part. It is shown that HMC considerably outperforms this approach in terms of effective sample size, MSE and observed coverage probabilities. In a second simulation study satisfactory performance is seen for the full HMC within Gibbs procedure. The approach is illustrated for a portfolio of financial assets with respect to one-day ahead value at risk forecasts. We provide comparison to a two-step estimation procedure of the proposed model and to relevant benchmark models: a model with dynamic linear models for the margins and a single factor copula for the dependence proposed by Schamberger et al(2017) and a multivariate factor stochastic volatility model proposed by Kastner et al(2017). Our proposed approach shows superior performance.

stat.CO↗

arXiv2019

A partial correlation vine based approach for modeling and forecasting multivariate volatility time-series

A novel approach for dynamic modeling and forecasting of realized covariance matrices is proposed. Realized variances and realized correlation matrices are jointly estimated. The one-to-one relationship between a positive definite correlation matrix and its associated set of partial correlations corresponding to any vine specification is used for data transformation. The model components therefore are realized variances as well as realized standard and partial correlations corresponding to a daily log-return series. As such, they have a clear practical interpretation. A method to select a regular vine structure, which allows for parsimonious time-series and dependence modeling of the model components, is introduced. Being algebraically independent the latter do not underlie any algebraic constraint. The proposed model approach is outlined in detail and motivated along with a real data example on six highly liquid stocks. The forecasting performance is evaluated both with respect to statistical precision and in the context of portfolio optimization. Comparisons with Cholesky decomposition based benchmark models support the excellent prediction ability of the proposed model approach.

stat.ME↗

arXiv2018

Model selection in sparse high-dimensional vine copula models with application to portfolio risk

Vine copulas allow to build flexible dependence models for an arbitrary number of variables using only bivariate building blocks. The number of parameters in a vine copula model increases quadratically with the dimension, which poses new challenges in high-dimensional applications. To alleviate the computational burden and risk of overfitting, we propose a modified Bayesian information criterion (BIC) tailored to sparse vine copula models. We show that the criterion can consistently distinguish between the true and alternative models under less stringent conditions than the classical BIC. The new criterion can be used to select the hyper-parameters of sparse model classes, such as truncated and thresholded vine copulas. We propose a computationally efficient implementation and illustrate the benefits of the new concepts with a case study where we model the dependence in a large stock stock portfolio.

stat.ME↗

arXiv2018

Vine copula based post-processing of ensemble forecasts for temperature

Today weather forecasting is conducted using numerical weather prediction (NWP) models, consisting of a set of differential equations describing the dynamics of the atmosphere. The output of such NWP models are single deterministic forecasts of future atmospheric states. To assess uncertainty in NWP forecasts so-called forecast ensembles are utilized. They are generated by employing a NWP model for distinct variants. However, as forecast ensembles are not able to capture the full amount of uncertainty in an NWP model, they often exhibit biases and dispersion errors. Therefore it has become common practise to employ statistical post processing models which correct for biases and improve calibration. We propose a novel post processing approach based on D-vine copulas, representing the predictive distribution by its quantiles. These models allow for much more general dependence structures than the state-of-the-art EMOS model and is highly data adapted. Our D-vine quantile regression approach shows excellent predictive performance in comparative studies of temperature forecasts over Europe with different forecast horizons based on the 52-member ensemble of the European Centre for Medium-Range Weather Forecasting (ECMWF). Specifically for larger forecast horizons the method clearly improves over the benchmark EMOS model.

stat.AP↗

arXiv2018

A Statistical Simulation Method for Joint Time Series of Non-stationary Hourly Wave Parameters

Statistically simulated time series of wave parameters are required for many coastal and offshore engineering applications, often at the resolution of approximately one hour. Various studies have relied on autoregressive moving-average (ARMA) processes to simulate synthetic series of wave parameters in a Monte Carlo sense. However, accurately representing inter-series dependencies has remained a challenge. In particular, the relationship between wave height and period statistics is complex, due to the limiting steepness condition. Here, we present a new simulation method for joint time series of significant wave height, mean zero-crossing periods and a directional regime variable. The latter distinguishes between northern and southwestern waves. The method rests on several model components which include renewal processes, Fourier series with random coefficients, ARMA processes, copulas and regime-switching. A particular feature is a data-driven estimate for a wave height-dependent limiting wave steepness condition which is used to facilitate copulabased dependence modeling. The method was developed for and applied to a data set in the Southern North Sea. For this site, the method could simulate time series with realistic annual cycles and inter-annual variability. In the time series data, the bivariate distribution of significant wave height and mean zero-crossing period was well represented. An influence of the directional regime on the bivariate distribution could also be modeled. However, the influence was not as strong in simulated data as in observed data. Finally, simulated series captured duration and inter-arrival time of storm events well. Potential applications for output of the simulation method range from the assessment of coastal risks or design of coastal structures to the planning and budgeting of offshore operations.

stat.AP↗

arXiv2018

Dependence modeling for recurrent event times subject to right-censoring with D-vine copulas

In many time-to-event studies, the event of interest is recurrent. Here, the data for each sample unit corresponds to a series of gap times between the subsequent events. Given a limited follow-up period, the last gap time might be right-censored. In contrast to classical analysis, gap times and censoring times cannot be assumed independent, i.e. the sequential nature of the data induces dependent censoring. Also, the recurrences typically vary between sample units leading to unbalanced data. To model the association pattern between gap times, so far only parametric margins combined with the restrictive class of Archimedean copulas have been considered. Here, taking the specific data features into account, we extend existing work in several directions: we allow for nonparametric margins and consider the flexible class of D-vine copulas. A global and sequential (one- and two-stage) likelihood approach are suggested. We discuss the computational efficiency of each estimation strategy. Extensive simulations show good finite sample performance of the proposed methodology. It is used to analyze the association in recurrent asthma attacks in children. The analysis reveals that a D-vine copula detects relevant insights, on how dependence changes in strength and type over time.

stat.ME↗

arXiv2017

Dependence Modeling in Ultra High Dimensions with Vine Copulas and the Graphical Lasso

To model high dimensional data, Gaussian methods are widely used since they remain tractable and yield parsimonious models by imposing strong assumptions on the data. Vine copulas are more flexible by combining arbitrary marginal distributions and (conditional) bivariate copulas. Yet, this adaptability is accompanied by sharply increasing computational effort as the dimension increases. The approach proposed in this paper overcomes this burden and makes the first step into ultra high dimensional non-Gaussian dependence modeling by using a divide-and-conquer approach. First, we apply Gaussian methods to split datasets into feasibly small subsets and second, apply parsimonious and flexible vine copulas thereon. Finally, we reconcile them into one joint model. We provide numerical results demonstrating the feasibility of our approach in moderate dimensions and showcase its ability to estimate ultra high dimensional non-Gaussian dependence models in thousands of dimensions.

stat.ML↗

arXiv2017

Vine copula based likelihood estimation of dependence patterns in multivariate event time data

In many studies multivariate event time data are generated from clusters having a possibly complex association pattern. Flexible models are needed to capture this dependence. Vine copulas serve this purpose. Inference methods for vine copulas are available for complete data. Event time data, however, are often subject to right-censoring. As a consequence, the existing inferential tools, e.g. likelihood estimation, need to be adapted. A two-stage estimation approach is proposed. First, the marginal distributions are modeled. Second, the dependence structure modeled by a vine copula is estimated via likelihood maximization. Due to the right-censoring single and double integrals show up in the copula likelihood expression such that numerical integration is needed for its evaluation. For the dependence modeling a sequential estimation approach that facilitates the computational challenges of the likelihood optimization is provided. A three-dimensional simulation study provides evidence for the good finite sample performance of the proposed method. Using four-dimensional mastitis data, it is shown how an appropriate vine copula model can be selected for data at hand.

stat.AP↗

arXiv2017

D-vine quantile regression with discrete variables

Quantile regression, the prediction of conditional quantiles, finds applications in various fields. Often, some or all of the variables are discrete. The authors propose two new quantile regression approaches to handle such mixed discrete-continuous data. Both of them generalize the continuous D-vine quantile regression, where the dependence between the response and the covariates is modeled by a parametric D-vine. D-vine quantile regression provides very flexible models, that enable accurate and fast predictions. Moreover, it automatically takes care of major issues of classical quantile regression, such as quantile crossing and interactions between the covariates. The first approach keeps the parametric estimation of the D-vines, but modifies the formulas to account for the discreteness. The second approach estimates the D-vine using continuous convolution to make the discrete variables continuous and then estimates the D-vine nonparametrically. A simulation study is presented examining for which scenarios the discrete-continuous D-vine quantile regression can provide superior prediction abilities. Lastly, the functionality of the two introduced methods is demonstrated by a real-world example predicting the number of bike rentals.

stat.ME↗

arXiv2017

A D-vine copula based model for repeated measurements extending linear mixed models with homogeneous correlation structure

We propose a model for unbalanced longitudinal data, where the univariate margins can be selected arbitrarily and the dependence structure is described with the help of a D-vine copula. We show that our approach is an extremely flexible extension of the widely used linear mixed model if the correlation is homogeneous over the considered individuals. As an alternative to joint maximum-likelihood a sequential estimation approach for the D-vine copula is provided and validated in a simulation study. The model can handle missing values without being forced to discard data. Since conditional distributions are known analytically, we easily make predictions for future events. For model selection we adjust the Bayesian information criterion to our situation. In an application to heart surgery data our model performs clearly better than competing linear mixed models.

stat.ME↗

arXiv2017

Selection of Sparse Vine Copulas in High Dimensions with the Lasso

We propose a novel structure selection method for high dimensional (d > 100) sparse vine copulas. Current sequential greedy approaches for structure selection require calculating spanning trees in hundreds of dimensions and fitting the pair copulas and their parameters iteratively throughout the structure selection process. Our method uses a connection between the vine and structural equation models (SEMs). The later can be estimated very fast using the Lasso, also in very high dimensions, to obtain sparse models. Thus, we obtain a structure estimate independently of the chosen pair copulas and parameters. Additionally, we define the novel concept of regularization paths for R-vine matrices. It relates sparsity of the vine copula model in terms of independence copulas to a penalization coefficient in the structural equation models. We illustrate our approach and provide many numerical examples. These include simulations and data applications in high dimensions, showing the superiority of our approach to other existing methods.

stat.ME↗

arXiv2017

Using model distances to investigate the simplifying assumption, model selection and truncation levels for vine copulas

Vine copulas are a useful statistical tool to describe the dependence structure between several random variables, especially when the number of variables is very large. When modeling data with vine copulas, one often is confronted with a set of candidate models out of which the best one is supposed to be selected. For example, this may arise in the context of non-simplified vine copulas, truncations of vines and other simplifications regarding pair-copula families or the vine structure. With the help of distance measures we develop a parametric bootstrap based testing procedure to decide between copulas from nested model classes. In addition we use distance measures to select among different candidate models. All commonly used distance measures, e.g. the Kullback-Leibler distance, suffer from the curse of dimensionality due to high-dimensional integrals. As a remedy for this problem, Killiches, Kraus and Czado (2017) propose several modifications of the Kullback-Leibler distance. We apply these distance measures to the above mentioned model selection problems and substantiate their usefulness.

stat.ME↗

arXiv2017

Stress Testing German Industry Sectors: Results from a Vine Copula Based Quantile Regression

Measuring interdependence between probabilities of default (PDs) in different industry sectors of an economy plays a crucial role in financial stress testing. Thereby, regression approaches may be employed to model the impact of stressed industry sectors as covariates on other response sectors. We identify vine copula based quantile regression as an eligible tool for conducting such stress tests as this method has good robustness properties, takes into account potential nonlinearities of conditional quantile functions and ensures that no quantile crossing effects occur. We illustrate its performance by a data set of sector specific PDs for the German economy. Empirical results are provided for a rough and a fine-grained industry sector classification scheme. Amongst others, we confirm that a stressed automobile industry has a severe impact on the German economy as a whole at different quantile levels whereas e.g., for a stressed financial sector the impact is rather moderate. Moreover, the vine copula based quantile regression approach is benchmarked against both classical linear quantile regression and expectile regression in order to illustrate its methodological effectiveness in the scenarios evaluated.

stat.AP↗

arXiv2017

Nonparametric estimation of simplified vine copula models: comparison of methods

In the last decade, simplified vine copula models have been an active area of research. They build a high dimensional probability density from the product of marginals densities and bivariate copula densities. Besides parametric models, several approaches to nonparametric estimation of vine copulas have been proposed. In this article, we extend these approaches and compare them in an extensive simulation study and a real data application. We identify several factors driving the relative performance of the estimators. The most important one is the strength of dependence. No method was found to be uniformly better than all others. Overall, the kernel estimators performed best, but do worse than penalized B-spline estimators when there is weak dependence and no tail dependence.

stat.ME↗