Purdue University

About: Purdue University is a education organization based out in West Lafayette, Indiana, United States. It is known for research contribution in the topics: Population & Context (language use). The organization has 73219 authors who have published 163563 publications receiving 5775236 citations. The organization is also known as: Purdue & Purdue-West Lafayette.

Ultrawide-Bandgap Semiconductors: Research Opportunities and Challenges

Abstract: J. Y. Tsao,* S. Chowdhury, M. A. Hollis,* D. Jena, N. M. Johnson, K. A. Jones, R. J. Kaplar,* S. Rajan, C. G. Van de Walle, E. Bellotti, C. L. Chua, R. Collazo, M. E. Coltrin, J. A. Cooper, K. R. Evans, S. Graham, T. A. Grotjohn, E. R. Heller, M. Higashiwaki, M. S. Islam, P. W. Juodawlkis, M. A. Khan, A. D. Koehler, J. H. Leach, U. K. Mishra, R. J. Nemanich, R. C. N. Pilawa-Podgurski, J. B. Shealy, Z. Sitar, M. J. Tadjer, A. F. Witulski, M. Wraback, and J. A. Simmons

Framing Sustainability in a Telecoupled World

Abstract: Interactions between distant places are increasingly widespread and influential, often leading to unexpected outcomes with profound implications for sustainability. Numerous sustainability studies have been conducted within a particular place with little attention to the impacts of distant interactions on sustainability in multiple places. Although distant forces have been studied, they are usually treated as exogenous variables and feedbacks have rarely been considered. To understand and integrate various distant interactions better, we propose an integrated framework based on telecoupling, an umbrella concept that refers to socioeconomic and environmental interactions over distances. The concept of telecoupling is a logical extension of research on coupled human and natural systems, in which interactions occur within particular geographic locations. The telecoupling framework contains five major interrelated components, i.e., coupled human and natural systems, flows, agents, causes, and effects. We illustrate the framework using two examples of distant interactions associated with trade of agricultural commodities and invasive species, highlight the implications of the framework, and discuss research needs and approaches to move research on telecouplings forward. The framework can help to analyze system components and their interrelationships, identify research gaps, detect hidden costs and untapped benefits, provide a useful means to incorporate feedbacks as well as trade-offs and synergies across multiple systems (sending, receiving, and spillover systems), and improve the understanding of distant interactions and the effectiveness of policies for socioeconomic and environmental sustainability from local to global levels.

Systematic measurements of identified particle spectra in pp, d+Au, and Au+Au collisions at the star detector.

Abstract: Identified charged-particle spectra of pi(+/-), K(+/-), p, and (p) over bar at midrapidity (vertical bar y vertical bar < 0.1) measured by the dE/dx method in the STAR (solenoidal tracker at the BNL Relativistic Heavy Ion Collider) time projection chamber are reported for pp and d + Au collisions at root s(NN) = 200 GeV and for Au + Au collisions at 62.4, 130, and 200 GeV. Average transverse momenta, total particle production, particle yield ratios, strangeness, and baryon production rates are investigated as a function of the collision system and centrality. The transverse momentum spectra are found to be flatter for heavy particles than for light particles in all collision systems; the effect is more prominent for more central collisions. The extracted average transverse momentum of each particle species follows a trend determined by the total charged-particle multiplicity density. The Bjorken energy density estimate is at least several GeV/fm(3) for a formation time less than 1 fm/c. A significantly larger net-baryon density and a stronger increase of the net-baryon density with centrality are found in Au + Au collisions at 62.4 GeV than at the two higher energies. Antibaryon production relative to total particle multiplicity is found to be constant over centrality, but increases with the collision energy. Strangeness production relative to total particle multiplicity is similar at the three measured RHIC energies. Relative strangeness production increases quickly with centrality in peripheral Au + Au collisions, to a value about 50% above the pp value, and remains rather constant in more central collisions. Bulk freeze-out properties are extracted from thermal equilibrium model and hydrodynamics-motivated blast-wave model fits to the data. Resonance decays are found to have little effect on the extracted kinetic freeze-out parameters because of the transverse momentum range of our measurements. The extracted chemical freeze-out temperature is constant, independent of collision system or centrality; its value is close to the predicted phase-transition temperature, suggesting that chemical freeze-out happens in the vicinity of hadronization and the chemical freeze-out temperature is universal despite the vastly different initial conditions in the collision systems. The extracted kinetic freeze-out temperature, while similar to the chemical freeze-out temperature in pp, d + Au, and peripheral Au + Au collisions, drops significantly with centrality in Au + Au collisions, whereas the extracted transverse radial flow velocity increases rapidly with centrality. There appears to be a prolonged period of particle elastic scatterings from chemical to kinetic freeze-out in central Au + Au collisions. The bulk properties extracted at chemical and kinetic freeze-out are observed to evolve smoothly over the measured energy range, collision systems, and collision centralities.

Automatic differentiation in machine learning: a survey

Abstract: Derivatives, mostly in the form of gradients and Hessians, are ubiquitous in machine learning. Automatic differentiation (AD), also called algorithmic differentiation or simply "autodiff", is a family of techniques similar to but more general than backpropagation for efficiently and accurately evaluating derivatives of numeric functions expressed as computer programs. AD is a small but established field with applications in areas including computational fluid dynamics, atmospheric sciences, and engineering design optimization. Until very recently, the fields of machine learning and AD have largely been unaware of each other and, in some cases, have independently discovered each other's results. Despite its relevance, general-purpose AD has been missing from the machine learning toolbox, a situation slowly changing with its ongoing adoption under the names "dynamic computational graphs" and "differentiable programming". We survey the intersection of AD and machine learning, cover applications where AD has direct relevance, and address the main implementation techniques. By precisely defining the main differentiation techniques and their interrelationships, we aim to bring clarity to the usage of the terms "autodiff", "automatic differentiation", and "symbolic differentiation" as these are encountered more and more in machine learning settings.

Oxides and nitrides as alternative plasmonic materials in the optical range [Invited]

Abstract: As alternatives to conventional metals, new plasmonic materials offer many advantages in the rapidly growing fields of plasmonics and metamaterials. These advantages include low intrinsic loss, semiconductor-based design, compatibility with standard nanofabrication processes, tunability, and others. Transparent conducting oxides such as Al:ZnO, Ga:ZnO and indium-tin-oxide (ITO) enable many high-performance metamaterial devices operating in the near-IR. Transition-metal nitrides such as TiN or ZrN can be substitutes for conventional metals in the visible frequencies. In this paper we provide the details of fabrication and characterization of these new materials and discuss their suitability for a number of metamaterial and plasmonic applications.