Showing papers by "University of Paderborn published in 2022"

Photoactive iron complexes: more sustainable, but still a challenge

Abstract: With the “Criticality Score” used as a benchmark for sustainability – potentials, strategies and challenges are discussed to replace noble metal compounds in photosensitizers by the sustainable alternative iron.

A Review of Digital Transformation on Supply Chain Process Management Using Text Mining

Abstract: Industry 4.0 technologies are causing a paradigm shift in supply chain process management. The digital transformation of the supply chains provides enormous benefits to organizations by empowering collaboration among multiple internal and external organizations and systems. This study presents a narrative review explaining the existing knowledge on digital transformation in supply chain process management using text mining. It summarizes the existing literature to explain the current state of the art in supply chain digitalization. This comprehensive review identifies the most important topics and technologies and determines the future trends in this emerging field. We investigate the articles published in Web of Science and Scopus databases and use text mining techniques (clustering and topic modeling) on the article contents. Using VOS viewer, a bibliometric analysis of 395 articles with 12,700 references is analyzed. The contents of the articles are explored using text mining approaches. The synthesized results reveal that the most important topics in digital transformation are “sustainable supply chain management” and “circular economy and industry 4.0 technologies”. The study further discovers big data, data analytics, blockchain, artificial intelligence, machine learning, and the Internet of Things as the most critical technologies for facilitating supply chain digital transformation. Finally, an overlay heatmap analysis of the research articles found that digital transformation, supply chain management, industry 4.0, decision-making, and sustainability are emerging trends in supply chain digitalization.

An integrated rough group multicriteria decision-making model for the ex-ante prioritization of infrastructure projects: The Serbian Railways case

Abstract: Railway transportation is the backbone of the economy significantly influences mobility and life quality in many developed and developing countries. Railway infrastructure investment problems are inherently complex and involve multiple and often conflicting criteria in an uncertain socio-economic environment. This study presents an integrated rough group FUll COnsistency Method (FUCOM) and Multi-Atributive Ideal-Real Comparative Analysis (MAIRCA) method for railway infrastructure project evaluation and prioritization. The FUCOM method evaluates the selection criteria through a simple algorithm, and the MAIRCA method prioritizes alternative projects through structured and systematic mathematical computations. Sensitivity analysis measures the impact of the criteria weights on the final results. Spearman's rank correlation coefficient assesses the effect of criteria weight variations on alternative rankings' stability. This study's main contribution is developing and implementing a comprehensive and robust framework for ex-ante evaluation and prioritization of railway infrastructure projects. A case study in Serbian Railways demonstrates the proposed integrated framework's applicability and efficacy in evaluating the railway infrastructure projects for Serbian Railways.

Symplectic integration of learned Hamiltonian systems

Abstract: Hamiltonian systems are differential equations that describe systems in classical mechanics, plasma physics, and sampling problems. They exhibit many structural properties, such as a lack of attractors and the presence of conservation laws. To predict Hamiltonian dynamics based on discrete trajectory observations, the incorporation of prior knowledge about Hamiltonian structure greatly improves predictions. This is typically done by learning the system's Hamiltonian and then integrating the Hamiltonian vector field with a symplectic integrator. For this, however, Hamiltonian data need to be approximated based on trajectory observations. Moreover, the numerical integrator introduces an additional discretization error. In this article, we show that an inverse modified Hamiltonian structure adapted to the geometric integrator can be learned directly from observations. A separate approximation step for the Hamiltonian data is avoided. The inverse modified data compensate for the discretization error such that the discretization error is eliminated. The technique is developed for Gaussian processes.

A private sustainable partner selection model for green public-private partnerships and regional economic development

Abstract: Many cities are struggling to keep pace with limited budgets and rapid growth. Economic development models involving public-private partnerships (P3s) can help drive economic revitalization. The choice of partners plays a vital role in the success or failure of sustainable P3 initiatives. In this study, we propose a novel integrated sustainable private partner selection framework in P3s. The proposed model is composed of the best-worst method (BWM), the weighted influence non-linear gauge system (WINGS), and the technique of order preference similarity to the ideal solution (TOPSIS). The BWM is used to identify the importance weights of the economic, environmental, social, and technological criteria. The WINGS method uses ideographic causal maps to analyze the intertwined criteria and their causal relations. TOPSIS is used to rank and select the private partners that will bring the “best value” to the partnership. We demonstrate the proposed method's applicability in a P3 initiative for sustainability, gentrification and neighborhood revitalization, and economic development in a northeastern US city. In this initiative, low density, low cost, and biodegradable agricultural waste and mushroom fibers are grown in vacant buildings to be used as a Styrofoam packaging replacement.

The role of learning strategies for performance in mathematics courses for engineers

Abstract: We analyse the predictive power of learning strategies for engineering students’ performance in mathematics. Learning strategies play an important role in self-regulated learning. Based on a new learning strategy questionnaire that takes into account the specifics of mathematical learning at universities, we investigated what were the strategies that correlate with performance and predict future performance. We present data of a longitudinal study with N = 361 engineering students regressing their performance on students’ use of their learning strategies as well as their prior performance. The results indicate that practicing but not repeating the content and resisting frustration predict students’ performance. We discuss the findings with a specific view on what is tested and why some elaboration strategies might not be rewarded in exams.

A parameter tuned hybrid algorithm for solving flow shop scheduling problems with parallel assembly stages

Abstract: In this paper, we study the scheduling problem for a customized production system consisting of a flow shop production line with a parallel assembly stage that produces various products in two stages. In the first stage of the production line, parts are produced using a flow shop production line, and in the second stage, products are assembled on one of the parallel assembly lines. The objective is to minimize the time required to complete all goods (makespan) using efficient scheduling. A mathematical model is developed; however, the model is NP-hard and cannot be solved in a reasonable amount of time. To solve this NP-hard problem, we propose two well-known metaheuristics and a hybrid algorithm. To calibrate and improve the performance of our algorithms, we employ the Taguchi method. We evaluate the performance of our hybrid algorithm with the two well-known methods of Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) and demonstrate that our hybrid algorithm outperforms both the GA and PSO approaches in terms of efficiency.

Spontaneous and explicit symmetry breaking of thermoacoustic eigenmodes in imperfect annular geometries

Abstract: This article deals with the symmetry breaking of azimuthal thermoacoustic modes in annular combustors. Using a nominally symmetric annular combustor, we present experimental evidence of a predicted spontaneous reflectional symmetry breaking, and also an unexpected explicit rotational symmetry breaking in the neighbourhood of the Hopf bifurcation which separates linearly stable azimuthal thermoacoustic modes from self-oscillating modes. We derive and solve a multidimensional Fokker–Planck equation to unravel a unified picture of the phase space topology. We demonstrate that symmetric probability density functions of the thermoacoustic state vector are elusive, because the effect of asymmetries, even imperceptible ones, is magnified close to the bifurcation. This conclusion implies that the thermoacoustic oscillations of azimuthal modes in real combustors will systematically exhibit a statistically dominant orientation of the mode in the vicinity of the Hopf bifurcation.

Regularity of Lie groups

Abstract: We solve the regularity problem for Milnor's infinite dimensional Lie groups in the $C^0$-topological context, and provide necessary and sufficient regularity conditions for the (standard) $C^k$-topological setting. We prove that the evolution map is $C^0$-continuous on its domain $\textit{iff}\hspace{1pt}$ the Lie group $G$ is locally $\mu$-convex. We furthermore show that if the evolution map is defined on all smooth curves, then $G$ is Mackey complete. Under the assumption that $G$ is locally $\mu$-convex, we show that each $C^k$-curve for $k\in \mathbb{N}_{\geq 1}\sqcup\{\mathrm{lip},\infty\}$ is integrable (contained in the domain of the evolution map) $\textit{iff}\hspace{1pt}$ $G$ is Mackey complete and $\mathrm{k}$-confined. The latter condition states that each $C^k$-curve in the Lie algebra $\mathfrak{g}$ of $G$ can be uniformly approximated by a special type of sequence that consists of piecewise integrable curves. A similar result is proven for the case $k\equiv 0$; and, we provide several mild conditions that ensure that $G$ is $\mathrm{k}$-confined for each $k\in \mathbb{N}\sqcup\{\mathrm{lip},\infty\}$. We finally discuss the differentiation of parameter-dependent integrals in the (standard) $C^k$-topological context. In particular, we show that if the evolution map is defined and continuous on $C^k([0,1],\mathfrak{g})$ for $k\in \mathbb{N}\sqcup\{\infty\}$, then it is smooth thereon $\textit{iff}\hspace{1pt}$ it is differentiable at zero $\textit{iff}\hspace{1pt}$ $\mathfrak{g}$ is $\hspace{0.2pt}$ Mackey$\hspace{1pt}/ \hspace{1pt}$integral$\hspace{1pt}$ complete for $k\in \mathbb{N}_{\geq 1}\sqcup\{\infty\}\hspace{1pt}/\hspace{1pt}k\equiv 0$. This result is obtained by calculating the directional derivatives explicitly, recovering the standard formulas that hold, e.g., in the Banach case.

Small-signal solutions of a two-dimensional doubly degenerate taxis system modeling bacterial motion in nutrient-poor environments

Abstract: The doubly degenerate nutrient taxis model u t = ∇ ⋅ ( u v ∇ u ) − ∇ ⋅ ( u 2 v ∇ v ) + l u v , x ∈ Ω , t > 0 , v t = Δ v − u v , x ∈ Ω , t > 0 , is considered in smoothly bounded convex subdomains of the plane, with l ≥ 0 . It is shown that for any p > 2 and each fixed nonnegative u 0 ∈ W 1 , ∞ ( Ω ) , a smallness condition exclusively involving v 0 can be identified as sufficient to ensure that an associated no-flux type initial–boundary value problem with ( u , v ) | t = 0 = ( u 0 , v 0 ) admits a global weak solution satisfying ess sup t > 0 ‖ u ( ⋅ , t ) ‖ L p ( Ω ) ∞ . The proof relies on the use of an apparently novel class of functional inequalities which provide estimates from below for certain Dirichlet integrals involving possibly degenerate weight functions.

TaintBench: Automatic real-world malware benchmarking of Android taint analyses

Abstract: Due to the lack of established real-world benchmark suites for static taint analyses of Android applications, evaluations of these analyses are often restricted and hard to compare. Even in evaluations that do use real-world apps, details about the ground truth in those apps are rarely documented, which makes it difficult to compare and reproduce the results. To push Android taint analysis research forward, this paper thus recommends criteria for constructing real-world benchmark suites for this specific domain, and presents TaintBench, the first real-world malware benchmark suite with documented taint flows. TaintBench benchmark apps include taint flows with complex structures, and addresses static challenges that are commonly agreed on by the community. Together with the TaintBench suite, we introduce the TaintBench framework, whose goal is to simplify real-world benchmarking of Android taint analyses. First, a usability test shows that the framework improves experts’ performance and perceived usability when documenting and inspecting taint flows. Second, experiments using TaintBench reveal new insights for the taint analysis tools Amandroid and FlowDroid: (i) They are less effective on real-world malware apps than on synthetic benchmark apps. (ii) Predefined lists of sources and sinks heavily impact the tools’ accuracy. (iii) Surprisingly, up-to-date versions of both tools are less accurate than their predecessors.

Small-signal solutions of a two-dimensional doubly degenerate taxis system modeling bacterial motion in nutrient-poor environments

Abstract: The doubly degenerate nutrient taxis model ut=∇⋅(uv∇u)−∇⋅(u2v∇v)+ℓuv,x∈Ω,t>0,vt=Δv−uv,x∈Ω,t>0,is considered in smoothly bounded convex subdomains of the plane, with ℓ≥0. It is shown that for any p>2 and each fixed nonnegative u0∈W1,∞(Ω), a smallness condition exclusively involving v0 can be identified as sufficient to ensure that an associated no-flux type initial–boundary value problem with (u,v)|t=0=(u0,v0) admits a global weak solution satisfying esssupt>0‖u(⋅,t)‖Lp(Ω)<∞. The proof relies on the use of an apparently novel class of functional inequalities which provide estimates from below for certain Dirichlet integrals involving possibly degenerate weight functions.

Laser-lithographically written micron-wide superconducting nanowire single-photon detectors

Abstract: Abstract We demonstrate the fabrication of micron-wide tungsten silicide superconducting nanowire single-photon detectors on a silicon substrate using laser lithography. We show saturated internal detection efficiencies with wire widths ranging from 0.59 µ m to 1.43 µ m under illumination at 1550 nm. We demonstrate both straight wires, as well as meandered structures. Single-photon sensitivity is shown in devices up to 4 mm in length. Laser-lithographically written devices allow for fast and easy structuring of large areas while maintaining a saturated internal efficiency for wire widths around 1 µ m.

Marginal College Wage Premiums Under Selection Into Employment

Abstract: Abstract We identify female long-term wage returns to college education using the educational expansion between 1960–90 in West Germany as exogenous variation for college enrolment. We estimate marginal treatment effects and propose a simple partial identification technique accounting for women selecting into employment due to having a college education. College-educated women are, on average, more than 18 percentage points more likely to be employed due to having a college education than those without college education. Taking this into account, we bound wage returns to 5.7%–13.9% per year of education completed (average treatment effects on the treated).

Learning Concept Lengths Accelerates Concept Learning in ALC

Abstract: Concept learning approaches based on refinement operators explore partially ordered solution spaces to compute concepts, which are used as binary classification models for individuals. However, the number of concepts explored by these approaches can grow to the millions for complex learning problems. This often leads to impractical runtimes. We propose to alleviate this problem by predicting the length of target concepts before the exploration of the solution space. By these means, we can prune the search space during concept learning. To achieve this goal, we compare four neural architectures and evaluate them on four benchmarks. Our evaluation results suggest that recurrent neural network architectures perform best at concept length prediction with a macro F-measure ranging from 38% to 92%. We then extend the CELOE algorithm, which learns ALC concepts, with our concept length predictor. Our extension yields the algorithm CLIP. In our experiments, CLIP is at least 7.5 $$\times $$ faster than other state-of-the-art concept learning algorithms for ALC—including CELOE—and achieves significant improvements in the F-measure of the concepts learned on 3 out of 4 datasets. For reproducibility, we provide our implementation in the public GitHub repository at https://github.com/dice-group/LearnALCLengths .

A parameter tuned hybrid algorithm for solving flow shop scheduling problems with parallel assembly stages

Abstract: In this paper, we study the scheduling problem for a customized production system consisting of a flow shop production line with a parallel assembly stage that produces various products in two stages. In the first stage of the production line, parts are produced using a flow shop production line, and in the second stage, products are assembled on one of the parallel assembly lines. The objective is to minimize the time required to complete all goods (makespan) using efficient scheduling. A mathematical model is developed; however, the model is NP-hard and cannot be solved in a reasonable amount of time. To solve this NP-hard problem, we propose two well-known metaheuristics and a hybrid algorithm. To calibrate and improve the performance of our algorithms, we employ the Taguchi method. We evaluate the performance of our hybrid algorithm with the two well-known methods of Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) and demonstrate that our hybrid algorithm outperforms both the GA and PSO approaches in terms of efficiency.