Suchergebnisse

Regressionsanalysen sind von Konfundierungseffekten betroffen, wenn Drittvariablen gleichzeitig mit Zielgrößen und Kovariablen korreliert sind. Klassische Regressionsmodelle sind in diesen Fällen nicht in der Lage, Kovariablen- und Drittvariableneffekte voneinander zu unterscheiden. Die resultierenden Schätzer sind folglich in der Regel verzerrt. Motivation für diese Dissertation ist der Spezialfall von räumlicher Konfundierung, der auftritt wenn Drittvariablen eine ähnliche räumliche Struktur in Kovariablen und Zielgrößen induzieren. Um der Verzerrung des geschätzten Kovariableneffektes en...

Regressionsanalysen sind von Konfundierungseffekten betroffen, wenn Drittvariablen gleichzeitig mit Zielgrößen und Kovariablen korreliert sind. Klassische Regressionsmodelle sind in diesen Fällen nicht in der Lage, Kovariablen- und Drittvariableneffekte voneinander zu unterscheiden. Die resultierenden Schätzer sind folglich in der Regel verzerrt. Motivation für diese Dissertation ist der Spezialfall von räumlicher Konfundierung, der auftritt wenn Drittvariablen eine ähnliche räumliche Struktur in Kovariablen und Zielgrößen induzieren. Um der Verzerrung des geschätzten Kovariableneffektes en...

High levels of carbon emissions and rising income inequality are interconnected challenges for the global society. Commonly-applied linear regression models fail to unravel the complexity of potential bi-directional transmission channels. Specifically, consumption, energy sources and the political system are potential determinants of the strength and direction of the dependence between emissions and inequality. To capture their impact, this study investigates the conditional dependence between income inequality and emissions by applying distributional copula models on an unbalanced panel data set of 154 countries from 1960 to 2019. A comparison of high-, middle-, and low-income countries contradicts a linear relationship and sheds light on heterogeneous dependence structures implying synergies, trade-offs and decoupling between income inequality and carbon emissions. Based on the conditional distribution, we can identify determinants associated with higher/lower probabilities of a country falling in an area of potential social and environmental sustainability.

We investigate a novel database of 10,217 extreme operational losses from the Italian bank UniCredit, covering a period of 10 years and 7 different event types. Our goal is to shed light on the dependence between the severity distribution of these losses and a set of macroeconomic, financial and firm-specific factors. To do so, we use Generalized Pareto regression techniques, where both the scale and shape parameters are assumed to be functions of these explanatory variables. In this complex distributional regression framework, we perform the selection of the relevant covariates with a state-of-the-art penalized-likelihood estimation procedure relying on $L_{1}$-norm penalty terms of the coefficients. A simulation study indicates that this approach efficiently selects covariates of interest and tackles spurious regression issues encountered when dealing with integrated time series. The results of our empirical analysis have important implications in terms of risk management and regulatory policy. In particular, we found that high Italian unemployment rate and low GDP growth rate in the European Union are associated with smaller probabilities of extreme severities, whereas high values of the VIX and high growth rates of housing prices are associated with more extreme losses. Looking at firm-specific factors, low leverage ratio and high deposit growth rate are associated with a higher likelihood of extreme losses. Lastly, we illustrate the impact of different economic scenarios on the requested capital for operational risk. We find important discrepancies across event types and scenarios. ; Peer reviewed

The effects of demographic change and lack of acceptance pose some of the main problems for public transport infrastructure in rural areas in developed countries. In developing countries, however, there is a need to optimize and increase the efficiency of transport. For both cases digitisation offers solutions that are investigated in this issertation?. For developing countries, simple IT applications were examined for possible efficiency improvements. For rural areas in Germany, the Max-Planck-Institute for Dynamics and Self-Organization has developed a new form of demand-oriented transpor...

The effects of demographic change and lack of acceptance pose some of the main problems for public transport infrastructure in rural areas in developed countries. In developing countries, however, there is a need to optimize and increase the efficiency of transport. For both cases digitisation offers solutions that are investigated in this issertation?. For developing countries, simple IT applications were examined for possible efficiency improvements. For rural areas in Germany, the Max-Planck-Institute for Dynamics and Self-Organization has developed a new form of demand-oriented transpor...

Urbanization is a global trend. Currently, about half of the world's population lives in cities. According to the latest calculations, this figure is expected to grow to 68 percent (about 6.8 billion people) by 2050. However, urbanization rates vary greatly from country to country and from continent to continent. Asia and Africa are the main hotspots of future urbanization, with India, China, and Nigeria alone accounting for 35 percent of urban population growth by 2050. At the same time, however, 90 percent of the world's rural population lives on these continents. Accordingly, there will ...

We reexamine the effect of economic development on the level of democ- racy based on the data sets of Acemoglu et al. (Am Econ Rev 98(3):808–842, 2008) with a novel regression specification utilizing a zero–one-inflated beta distribution for the response variable democracy. Contrary to the results of Acemoglu et al. ( 2008), some support for a positive association between income and democracy is found when assuming that the variance of explanatory variables is heterogenous. In particular, our results show that rising income is associated with a probability of becoming fully democratic, but income is not generally associated with the mean level of democracy

We reexamine the effect of economic development on the level of democracy based on the data sets of Acemoglu et al. (2008) with a novel regression specification utilizing a zero-one-inflated beta distribution for the response variable democracy. The zero-one-inflated beta distribution is more appropriate for continuous but bounded responses with non-zero probabilities for the boundaries of the support than the other frequently used distributions such as the normal. Contrary to the results of Acemoglu et al. (2008), some support of causality is found particularly when explaining the variance of the democracy variables. Since our analysis highlights that the distribution of democracy is bimodal, we approximate the modes using two separate samples of OECD and non-OECD countries. Our results indicate that there are differences not only in the mean but also in other features of the response distribution between the two groups. For instance, higher incomes are associated with higher democracy levels in the OECD sub-sample, however for non-OECD the association is insignificant.

The aim of this book is an applied and unified introduction into parametric, non- and semiparametric regression that closes the gap between theory and application. The most important models and methods in regression are presented on a solid formal basis, and their appropriate application is shown through many real data examples and case studies. Availability of (user-friendly) software has been a major criterion for the methods selected and presented. Thus, the book primarily targets an audience that includes students, teachers and practitioners in social, economic, and life sciences, as well as students and teachers in statistics programs, and mathematicians and computer scientists with interests in statistical modeling and data analysis. It is written on an intermediate mathematical level and assumes only knowledge of basic probability, calculus, and statistics. The most important definitions and statements are concisely summarized in boxes. Two appendices describe required matrix algebra, as well as elements of probability calculus and statistical inference

AbstractResponse functions that link regression predictors to properties of the response distribution are fundamental components in many statistical models. However, the choice of these functions is typically based on the domain of the modeled quantities and is usually not further scrutinized. For example, the exponential response function is often assumed for parameters restricted to be positive, although it implies a multiplicative model, which is not necessarily desirable or adequate. Consequently, applied researchers might face misleading results when relying on such defaults. For parameters restricted to be positive, we propose to construct alternative response functions based on the softplus function. These response functions are differentiable and correspond closely to the identity function for positive values of the regression predictor implying a quasi-additive model. Consequently, the proposed response functions allow for an additive interpretation of the estimated effects by practitioners and can be a better fit in certain data situations. We study the properties of the newly constructed response functions and demonstrate the applicability in the context of count data regression and Bayesian distributional regression. We contrast our approach to the commonly used exponential response function.

Die aus dem maschinellen Lernen hervorgegangene statistische Boostingtechnik ist zu einer weit verbreiteten Methode zur Schätzung statistischer Modelle geworden. Als eine der erfolgreichsten Varianten wird das komponentenweise Gradientenboosting von immer mehr Statistikern favorisiert, da das iteratives Verfahren nicht nur eine intuitive Variablenauswahl in der Analyse von hochdimensionalen Datensätzen ermöglicht, sondern auch zusätzliche Flexibilität bietet, um verschiedene Arten von additiven Regressionstermen zu schätzen. Da Boostingmodelle nur Punktschätzer liefern, lassen sich in der R...