17315 research outputs found
Sort by
Total factor productivity and spillover effects: Frontier and Laggard firms dynamics
Full text linkIn this paper, I explore the spillover effects of frontier firms on other firms in Türkiye, using a detailed administrative dataset with firm-level data on balance sheets, inter-firm transactions, and employment. I review key production function estimators, evaluate their assumptions and performance using a large dataset of Turkish firms, and use productivity estimates to identify frontier firms and assess their influence on laggard firms' performance. Additionally, I contribute to the empirical literature by exploring the spillover and network effects of frontier firms on laggard firms, as well as examining the productivity convergence of laggard firms to frontier firms. The analysis reveals three key findings: (i) Frontier firms generate positive spillover effects within sectors, which enhance sales, employment, exports, and asset growth among laggard firms; (ii) detailed firm-to-firm invoice data reveals that a higher share of frontier firms in a firm's network significantly boosts investment, net sales, and productivity growth; and (iii) laggard firms show faster productivity growth, with substantial variation across firm types and industries
FET modeling with deep neural networks and GAN-augmented small measurement dataset
No full textWe present FET modeling using Deep Neural Networks with a small but augmented dataset. NMOS-HV measurement data from the IHP SG13G2 OpenPDK repository, including FET dimensions, VGS, VDS, and VBS, were used for DC drain current prediction. To improve accuracy, we employed Generative Adversarial Networks (GANs) for data augmentation. Using only 78,972 measured data points with 80-20% training-test split, we achieved 11.16% MAPE across 27 FET sizes. Using DL Model on GAN-based augmented dataset further reduced the error to 9.84%, demonstrating that accurate FET models can be developed with significantly fewer data points than traditional ANN approaches. Our results highlight the potential of data-efficient Deep Learning and GAN augmentation for automated transistor modeling
Acoustic characterization of cavitating flows in a microfluidic venturi reactor
No full textThis study investigates cavitating flows within a silicon-glass microfluidic venturi reactor using an integrated multi-modal approach that combines cavitation-induced noise analysis and high-speed imaging. Acoustic signal processing, primarily via power spectral density (PSD) and time-resolved frequency analysis (using wavelet transform), was employed to characterize cavitating flows across a wide range of upstream pressures (0.76 to 4.52 MPa). To the best knowledge of the authors, this is the first study on the direct correlation of time-resolved cavitating flow morphology captured via high-speed imaging of arrival, growth, and shedding of cavities with their acoustic signatures in both PSD diagrams and wavelet spectrograms in micro scale. By mapping specific PSD peaks and time-localized wavelet bursts to distinct flow regimes, a unified framework spanning inception through fully developed cavitation was constructed for a micro-venturi reactor. At lower pressures (<0.76 MPa), acoustic emissions in the 2-8 kHz range corresponded to cavitation inception with low cavitation intensity. As the upstream pressure was increased, a shift toward higher frequencies (up to and beyond 18 kHz) and a broadening of the spectral distribution revealed a transition to more intense sheet and cloud cavitation regimes. These acoustic signatures were confirmed by visual observations, which captured the evolution of distinct cavitating flow patterns. The findings highlight the potential of acoustic noise measurements as a real-time diagnostic tool for optimizing microscale hydrodynamic cavitation processes, with high potential for their implementation in applications such as wastewater treatment, energy harvesting, biomedical engineering, and integrated lab-on-a-chip platforms
Componentwise linearity of dominating ideals of path graphs
Full text linkWe show the componentwise linearity of dominating ideals of path graphs by describing a linear quotient order of their minimal generating sets. We also give formulas for their Betti numbers, regularity and projective dimension
Harnessing Graphene's potential for polymer electrolyte membrane fuel cells: challenges and future horizons
Full text linkInnovative deposition strategies for carbon-supported Pt catalysts in electrochemical hydrodeoxygenation of biomass-derived oxygenate
Full text linkA comparative study of multi-label supervised contrastive losses for the content-based image retrieval of remote sensing images
No full textContrastive loss has been extensively studied for both supervised and unsupervised learning, and its success has led to its extension towards multi-label classification scenarios. Its application to multi-label content-based image retrieval however has been very limited. This study provides a systematic comparison in the context of content based remote sensing image retrieval where ever-growing data collections require efficient and effective management tools. Four multi-label supervised contrastive losses were investigated along with three benchmark datasets: BigEarthNet, UC Merced, and ML-AID. The weighted MulSupCon method achieved up to 90.11% mAP on ML-AID and 70.32% mAP on UC Merced, demonstrating its effectiveness in multi-label remote sensing retrieval tasks
Symmetric informationally complete positive operators-valued measures and the Knaster’s conjecture
No full textSymmetric Informationally Complete Positive Operator-Valued Measures (SIC-POVMs) have been constructed in many dimensions using the Weyl-Heisenberg group. In the quantum information community, it is commonly believed that SIC-POVMs exist in all dimensions; however, the general proof of their existence is still an open problem. By mapping SIC-POVMs onto the generalized Bloch sphere, we prove two geometric existence statements associated with the SIC-POVM existence problem. First, we prove the Knaster’s conjecture for n vertices of an n-simplex and use that to prove the existence of a continuous family of general SIC-POVMs where the trace of the kth power of the operators is the same for (n2 − 1) of the elements. Furthermore, by using numerical methods, we show that in dimensions 3 and 4, a regular simplex can be constructed on the Bloch sphere such that all its vertices correspond to density matrices having the same trace of ρ3. In the three-dimensional Hilbert space, we generate 104 general SIC-POVMs for where all the elements have the same set of randomly chosen eigenvalues, indicating the continuous set of general SIC-POVMs can be constructed for all possible values of trace of ρ
Clean rides, healthy lives: the impact of electric vehicle adoption on air quality and infant health
No full textThis paper provides the first nationwide evidence on how electric vehicle (EV) adoption has improved both air quality and child health. We assemble a rich dataset from 2010–2021 that links county-level EV registrations to measures of air pollution, birth outcomes, and emergency department visits. The endogeneity of EV adoption is addressed using two complementary strategies: Two-way fixed effects and instrumental variables (IV). The IV exploits the staggered rollout of Alternative Fuel Corridors as a source of exogenous variation in charging infrastructure that affected EV adoption. The estimates show that greater EV penetration significantly reduces nitrogen dioxide (NO2), a key pollutant linked to vehicular emissions. These improvements in air quality yield significant health benefits, including reductions in very low birth weight and very premature births, as well as fewer asthma-related emergency department visits among children ages 0 to 5. This is true even when potentially offsetting increases in pollution from the electricity generation needed to power EVs are accounted for. The benefits are higher in the high-pollution counties with Alternative Fuel Corridors, where baseline exposures are greatest. The resulting reductions in very low birth weight births alone could generate annual benefits of 4.0 billion. These findings underscore the dual environmental and public health benefits of EV adoption
Functional island scanning strategy for powder bed fusion
Full text linkEnergy is transferred from the energy beam to the processed material during additive manufacturing (AM). The high energy input and uneven temperature distribution can result in a substantial temperature gradient, leading to significant thermal stress and warping deformation. One of the most influential factors in the material processing of additive manufacturing is the choice of scanning strategies. Therefore, optimizing the scanning strategy can help achieve a more even heat distribution across the part and reduce thermal distortion, which is crucial for producing high-quality SLM components. In this study, we developed a modified island scanning strategy and thermal management approach which is related to ordering the scanning sequence of islands, similar to the Traveling Salesman Problem (TSP) variant and it is also adapted for multiple laser cases. The introduction of modified island scan strategies aims to preserve the geometric integrity of the part, minimize distortion, and optimize the scanning process by considering the part’s geometric features, all while achieving the desired mechanical properties