In engineering, the modeling of complex systems plays a central role. Increasing computing power and storage capacities as well as the trend towards deep neural networks are resulting in more and more data being stored. This dissertation addresses two main challenges arising from handling large amounts of data for data-driven modeling: Firstly, the choice of a subset that is representative of the dataset from which it is selected. Secondly, the handling of unbalanced datasets, i.e., datasets with regimes of higher and lower point density. The first challenge is addressed by developing a novel subset selection algorithm based on kernel density estimation. The method ensures that the selected subset is representative of the original dataset or any desired arbitrary distribution. A sophisticated yet simple approach to evaluating the estimated density allows to save computing time. The second challenge is addressed by introducing a data weighting method that extends the standard loss function. The weights for the individual data points are adjusted in such a way that data points from regions of sparser point densities are weighted higher and data points from regions of higher point densities are weighted lower in order to ensure a more balanced model performance. This approach is independent of model architecture and suited for any training algorithm. The effectiveness of the developed methods is demonstrated by using benchmark datasets and real-world application examples. Among others, the examples of thermal modeling of a permanent magnet synchronous motor and a cold forming process are used. The results show that the presented method for subset selection can effectively select representative datasets and is on par with state-of-the-art approaches to subset selection. Additionally, the presented method is able to select the subset to represent arbitrary desired pdfs which gives the user much freedom of design. The introduced method for data weighting typically results in significant performance improvements for dynamic models, especially for imbalanced training datasets. Overall, these contributions provide a valuable contribution to the further development of data-driven modeling methods and offer practicable solutions for real-world challenges.

The final report of the DFG-funded project ‘State History on the Radio: The Production Process and the intermedial Transfer of Regional Historical Narratives by the WDR Regional Editorial Office under Walter Först (1960–early 1990s)’ briefly summarises the results of the historical research. Using North Rhine-Westphalia as an example, the project analyses narratives of state history distributed by mass media between 1960 and the early 1990s and traces practices of "doing history". In a symbiosis of public history and media and state historical perspective, the intermedial transfer of historical contributions as building modules of historical culture will be examined exemplarily on the basis of the WDR-Hörfunk-Landesredaktion (WDR radio editorial office dealing with state affairs) around the journalistic NRW history expert Walter Först (1920-1993). The editorial team has actively produced and anchored formative historical images for North Rhine-Westphalia. On the basis of intensive archival research, it analyses how the journalistic actors located the state in the field of tension of a diverse region, the Federal Republic of Germany and its European neighbours and how they promoted historical processes of meaning and identity formation. It is examined how the editor-in-chief established a network with other institutions and persons working in the fields of politics, culture and history, transferred this into institutional forms beyond broadcasting and thus created an additional communication basis for negotiated narratives.

Russia’s war in Ukraine has been marked by both parties’ extensive use of drones. While the early phases of the war saw the development of what we have called Consumer Drone Warfare, that is, the use of off-the-shelf or ready-to-fly drones provided by large manufacturers like DJI or Autel, there has been a notable shift towards locally assembled First-Person-View or FPV Drones in recent times. We argue that this development is made possible by – and in turn fuels – the platformization of warfare by the Ukrainian military. In this article, we trace three distinct but interwoven sites of platformization: 1) the modular FPV drone itself, understood as a military platform for sensor and weapon systems; 2) the Delta System, an ecosystem of applications for planning and cooperatively executing drone strikes; 3) and the Brave1 Market, a logistics platform connecting military units to equipment and weapon manufacturers which has been dubbed “the Amazon for Drone Warfare.” In doing so, we will show that platformized drone warfare reshapes and economizes processes of military organization and procurement according to the logics of markets and gamification. As they render the chain of operations from drone development to battlefield deployment (ac)countable, Delta and Brave1 are designed to facilitate quality control and self-organization vis-a-vis an increasing number of military tech start-ups and heterogenous hardware parts. We posit that the infrastructures for drone development and data collection established by platformized FPV drone warfare act as stepping stones towards (semi-)autonomous forms of drone warfare powered by machine learning and AI systems.

This thesis consists of five thematically related chapters on inequality, digitalization, and the environment. Two recurring themes are (i) who captures rents under shocks and how institutions shape income distribution, and (ii) how inequality and digital technologies alter emissions with policy-relevant trade-offs. The first chapter decomposes Germany’s 2021-2023 inflation across 11 sectors using Gross Value Added deflators, showing that profits - not wages - captured 57.6% of price growth in high-inflation sectors, with implications for competition and fiscal policy to countermeasure inflation and distributional impacts. Chapter 2 constructs Ecuador’s first Distributional National Accounts (1990-2022) from survey-tax microdata, finding persistent top-end concentration (top one percent: 25% of pretax income) and a fragile move from extractive to weakly inclusive institutions during the commodity boom that eroded post-COVID. Chapter 3 integrates tax registers and machine learning-based consumption to account for the carbon emissions of the "missing rich": the top ten percent share of disposable income rises from 30% (survey) to 46% (integrated) and their emissions share from 26% to 31%. However, full redistribution would raise emissions by 26% versus 6% in survey-only estimates - highlighting needs for the design of emission-mitigating in combination with income redistributional policies. Chapter 4 estimates the net climate effect of digitalization across industrialized countries: digitalization in firms and households generally lowers carbon dioxide emissions, but the optimum differs - lower-income industrialized countries lie above its potential emission-reducing level of digitalization while higher-income ones remain below efficiency-enhancing levels. Chapter 5 investigates digitalization in the agricultural sector: adopting digital yield monitoring in 28 states in the United States (1996-2010) reduces nitrous oxide per unit of output and per hectare; production expands, yet efficiency gains dominate, delivering net emission cuts and provide evidence for climate-smart agriculture.