Search NASA⌕ Search

SEARCH · Search NASA

Results for “classifiers”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 739 records · Page 41
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Astronomical image time series classification using CONVolutional attENTION (ConvEntion)

Aims: The treatment of astronomical image time series has won increasing attention in recent years. Indeed, numerous surveys following up on transient objects are in progress or under construction, such as the Vera C. Rubin Observatory Legacy Survey for Space and Time (LSST), which is poised to produce huge amounts of these time series. The associated scientific topics are extensive, ranging from the study of objects in our galaxy to the observation of the most distant supernovae for measuring the expansion of the universe. With such a large amount of data available, the need for robust automatic tools to detect and classify celestial objects is growing steadily. Methods: This study is based on the assumption that astronomical images contain more information than light curves. In this paper, we propose a novel approach based on deep learning for classifying different types of space objects directly using images. We named our approach ConvEntion, which stands for CONVolutional attENTION. It is based on convolutions and transformers, which are new approaches for the treatment of astronomical image time series. Our solution integrates spatio-temporal features and can be applied to various types of image datasets with any number of bands. Results: In this work, we solved various problems the datasets tend to suffer from and we present new results for classifications using astronomical image time series with an increase in accuracy of 13%, compared to state-of-the-art approaches that use image time series, and a 12% increase, compared to approaches that use light curves.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2024

Barium stars as tracers of s -process nucleosynthesis in AGB stars

Barium (Ba) stars help to verify asymptotic giant branch (AGB) star nucleosynthesis models since they experienced pollution from an AGB binary companion and thus their spectra carry the signatures of the slow neutron capture process (s process). For a large number (180) of Ba stars, we searched for AGB stellar models that match the observed abundance patterns. We aim to uncover any systematic deviations of the sample abundances from the predictions of the nucleosynthesis models. We employed three machine learning algorithms as classifiers: a Random Forest method, developed for this work, and the two classifiers used in our previous study. Compared to that work, we also expanded our observational sample with 11 Ba stars available in the supersolar metallicity range. We studied the statistical behaviour of the different s-process elements in the observational sample to investigate if the AGB models systematically under- or overpredict the abundances observed in the Ba stars and show the results in the form of violin plots of the residuals between spectroscopic abundances and model predictions. We inspected the correlations between the observed [Fe/H], the s-process elemental abundances, and the residuals. We employed the [Zr/Fe] and [Nb/Fe] abundances as a thermometer to constrain the operational temperature that rules the production of these elements in the sample stars, assuming a steady-state s process. We also investigated the mass distribution of the identified polluter AGB stars and the behaviour of the δ parameter, which describes the fraction of accreted AGB material relative to the Ba star envelope. We find a significant trend in the residuals that implies an underproduction of the elements just after the first s-process peak (Nb, Mo, and Ru) in the models relative to the observations. This may originate from a neutron-capture process (e.g. the intermediate neutron-capture process, i process) not yet included in the AGB models of metallicity from solar to roughly 1/5 solar, corresponding to the range of the Ba stars. Correlations are found between the residuals of these peculiar elements, suggesting a common origin for the deviations from the models. In addition, there is a weak metallicity dependence of the residuals of these elements. The s-process temperatures derived with the [Zr/Fe] – [Nb/Fe] thermometer have an unrealistic value for the majority of our stars. The most likely explanation is that at least a fraction of these elements are not produced in a steady-state s process, and instead may be due to processes not included in the AGB models. The mass distribution of the identified models confirms that our sample of Ba stars was polluted by low-mass AGB stars (< 4 M ⊙ ). Most of the matching AGB models require low accreted mass, but a few systems with high accreted mass are needed to explain the observations.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2021

Graph Neural Network for Object Reconstruction in Liquid Argon Time Projection Chambers

This paper presents a graph neural network (GNN) technique for low-level reconstruction of neutrino interactions in a Liquid Argon Time Projection Chamber (LArTPC). GNNs are still a relatively novel technique, and have shown great promise for similar reconstruction tasks in the LHC. In this paper, a multihead attention message passing network is used to classify the relationship between detector hits by labelling graph edges, determining whether hits were produced by the same underlying particle, and if so, the particle type. The trained model is 84% accurate overall, and performs best on the EM shower and muon track classes. The model’s strengths and weaknesses are discussed, and plans for developing this technique further are summarised.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2022

Identifying Modular Construction Worker Tasks Using Computer Vision

Modular construction is increasingly being seen as an attractive method for delivering building projects due to advantages in safety, quality, and lead-time. Despite these benefits, this method still relies heavily on human labor, which causes variability in factory assembly-line performance that can erode performance benefits of modular construction. Continuous improvement methods can alleviate some of these issues, but they also require continuous monitoring of human workers' performance. Due to limitations of manual time study and automated sensor-based monitoring methods, recently computer vision-based methods have gained momentum in identifying the activities of construction workers from the videos of onsite construction. Therefore, this paper explores the use of computer vision-based human activity recognition techniques to identify and classify worker activities in modular construction videos. Computer vision-based tracking method has been used to track the human workers in each frame, and Resnet-50 network has been used to classify the activity of tracked workers. Evaluation of this framework has achieved higher than 90% accuracy and recall in testing.

computer vision↗
DOE OSTI2014

Binding energetics of substitutional and interstitial helium and di-helium defects with grain boundary structure in α-Fe

The formation/binding energetics and length scales associated with the interaction between He atoms and grain boundaries in BCC α-Fe were explored. Ten different low Σ grain boundaries from the 〈100〉 and 〈110〉 symmetric tilt grain boundary systems were used. In this work, we then calculated formation/binding energies for 1–2 He atoms in the substitutional and interstitial sites (HeV, He{sub 2}V, HeInt, He{sub 2}Int) at all potential grain boundary sites within 15 Å of the boundary (52 826 simulations total). The present results provide detailed information about the interaction energies and length scales of 1–2 He atoms with grain boundaries for the structures examined. A number of interesting new findings emerge from the present study. For instance, the Σ3(112) twin boundary in BCC Fe possesses a much smaller binding energy than other boundaries, which corresponds in long time dynamics simulations to the ability of an interstitial He defect to break away from the boundary in simulations on the order of nanoseconds. Additionally, positive correlations between the calculated formation/binding energies of the He defects (R > 0.9) asserts that the local environment surrounding each site strongly influences the He defect energies and that highly accurate quantum mechanics calculations of lower order defects may be an adequate predictor of higher order defects. Various metrics to quantify or classify the local environment were compared with the He defect binding energies. The present work shows that the binding and formation energies for He defects are important for understanding the physics of He diffusion and trapping by grain boundaries, which can be important for modeling He interactions in polycrystalline steels.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2014

General displaced SU(1, 1) number states: Revisited

The most general displaced number states, based on the bosonic and an irreducible representation of the Lie algebra symmetry of su(1, 1) and associated with the Calogero-Sutherland model are introduced. Here, we utilize the Barut-Girardello displacement operator instead of the Klauder-Perelomov counterpart, to construct new kind of the displaced number states which can be classified in nonlinear coherent states regime, too, with special nonlinearity functions. They depend on two parameters, and can be converted into the well-known Barut-Girardello coherent and number states, respectively, depending on which of the parameters equal to zero. A discussion of the statistical properties of these states is included. Significant are their squeezing properties and anti-bunching effects which can be raised by increasing the energy quantum number. Depending on the particular choice of the parameters of the above scenario, we are able to determine the status of compliance with flexible statistics. Major parts of the issue is spent on something that these states, in fact, should be considered as new kind of photon-added coherent states, too. Which can be reproduced through an iterated action of a creation operator on new nonlinear Barut-Girardello coherent states. Where the latter carry, also, outstanding statistical features.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2014

Spall behavior of cast iron with varying microstructures

The spall strength of cast iron with varying microstructures has been investigated using plate impact at moderate speed. Stress history measurements were made with manganin stress gauges embedded between the back face of the specimen and a low impedance polycarbonate backing. Five separate cast irons were tested. Four of these consisted of gray cast iron with graphite in flake form, with three classified as Type VII A2 and the fourth containing a bimodal distribution of Types VII A4 and VII D8. The fifth casting consisted of ductile cast iron with graphite in nodular form, classified as Type I, size class 5. The spall strength for the Type VII A2 gray cast irons varied between 40 and 370 MPa, and that of the additional gray cast iron, between 410 and 490 MPa. The spall strength of the ductile cast iron fell within the range of 0.94–1.2 GPa. It is shown that the spall strength is linked to the damage level at the spall plane, where an increased level of tensile stress is required to generate higher levels of damage. Post mortem analysis was performed on the recovered samples, revealing the graphite phase to be the primary factor governing the spall fracture of cast irons, where crack nucleation is directly correlated to the debonding of graphite from the metal matrix. The average length of graphite found within a casting is linked to the material's strength, where strength increases as a function of decreasing length. The morphology and mean free path of graphite precipitates further govern the subsequent coalescence of initiated cracks to form a complete fracture plane. In cases where graphite spacing is large, increased energy level is required to complete the fracture process. A secondary factor governing the spall fracture of cast irons has also been linked to the microstructure of the metal matrix, with pearlite yielding higher spall strengths than free ferrite.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2014

A parametric study of the drift-tearing mode using an extended-magnetohydrodynamic model

The linear, collisional, constant-ψ drift-tearing mode is analyzed for different regimes of the plasma-β, ion-skin-depth parameter space with an unreduced, extended-magnetohydrodynamic model. Here, new dispersion relations are found at moderate plasma β and previous drift-tearing results are classified as applicable at small plasma β.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2015

An analysis of hydrated proton diffusion in ab initio molecular dynamics

A detailed understanding of the inherently multiscale proton transport process raises a number of scientifically challenging questions. For example, there remain many (partially addressed) questions on the molecular mechanism for long-range proton migration and the potential for the formation of long-lived traps giving rise to burst-and-rest proton dynamics. Using results from a sizeable collection of ab initio molecular dynamics (AIMD) simulations (totaling ∼2.7 ns) with various density functional approximations (Becke-Lee-Yang-Parr (BLYP), BLYP–D3, Hamprecht-Cohen-Tozer-Handy, B3LYP) and temperatures (300–330 K), equilibrium and dynamical properties of one excess proton and 128 water molecules are studied. Two features in particular (concerted hops and weak hydrogen-bond donors) are investigated to identify modes in the system that are strongly correlated with the onset of periods of burst-and-rest dynamics. The question of concerted hops seeks to identify those time scales over which long-range proton transport can be classified as a series of sequential water hopping events or as a near-simultaneous concerted process along compressed water wires. The coupling of the observed burst-and-rest dynamics with motions of a fourth neighboring water molecule (a weak hydrogen-bond donor) solvating the protonated water molecule is also investigated. The presence (absence) of hydrogen bonds involving this fourth water molecule before and after successful proton hopping events is found to be strongly correlated with periods of burst (rest) dynamics (and consistent with pre-solvation concepts). By analyzing several realizations of the AIMD trajectories on the 100-ps time scale, convergence of statistics can be assessed. For instance, it was observed that the probability for a fourth water molecule to approach the hydronium, if not already proximal at the beginning of the lifetime of the hydronium, is very low, indicative of the formation of stable void regions. Furthermore, the correlations of the neighboring water atoms are identified as the fourth water approaches the hydronium. Finally, the temperature effects on structural and dynamical properties are studied.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2015

Impact of implanted phosphorus on the diffusivity of boron and its applicability to silicon solar cells

Boron diffusivity reduction in extrinsically doped silicon was investigated in the context of a process combination consisting of BBr{sub 3} furnace diffusion and preceding Phosphorus ion implantation. The implantation of Phosphorus leads to a substantial blocking of Boron during the subsequent Boron diffusion. First, the influences of ion implantation induced point defects as well as the initial P doping on B diffusivity were studied independently. Here, it was found that not the defects created during ion implantation but the P doping itself results in the observed B diffusion retardation. The influence of the initial P concentration was investigated in more detail by varying the P implantation dose. A secondary ion mass spectrometry (SIMS) analysis of the BSG layer after the B diffusion revealed that the B diffusion retardation is not due to potential P content in the BSG layer but rather caused by the n-type doping of the crystalline silicon itself. Based on the observations the B diffusion retardation was classified into three groups: (i) no reduction of B diffusivity, (ii) reduced B diffusivity, and (iii) blocking of the B diffusion. The retardation of B diffusion can well be explained by the phosphorus doping level resulting in a Fermi level shift and pairing of B and P ions, both reducing the B diffusivity. Besides these main influences, there are probably additional transient phenomena responsible for the blocking of boron. Those might be an interstitial transport mechanism caused by P diffusion that reduces interstitial concentration at the surface or the silicon/BSG interface shift due to oxidation during the BBr{sub 3} diffusion process. Lifetime measurements revealed that the residual (non-blocked) B leads to an increased dark saturation current density in the P doped region. Nevertheless, electrical quality is on a high level and was further increased by reducing the B dose as well as by removing the first few nanometers of the silicon surface after the BBr{sub 3} diffusion.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2015

Earthquakes clustering based on the magnitude and the depths in Molluca Province

In this paper, we present a model to classify the earthquakes occurred in Molluca Province. We use K-Means clustering method to classify the earthquake based on the magnitude and the depth of the earthquake. The result can be used for disaster mitigation and for designing evacuation route in Molluca Province.

58 GEOSCIENCES↗
DOE OSTI2016

Broadband angular selectivity of light at the nanoscale: Progress, applications, and outlook

Humankind has long endeavored to control the propagation direction of light. Since time immemorial, shades, lenses, and mirrors have been used to control the flow of light. In modern society, with the rapid development of nanotechnology, the control of light is moving toward devices at micrometer and even nanometer scales. At such scales, traditional devices based on geometrical optics reach their fundamental diffraction limits and cease to work. Nano-photonics, on the other hand, has attracted wide attention from researchers, especially in the last decade, due to its ability to manipulate light at the nanoscale. This review focuses on the nano-photonics systems that aim to select light based on its propagation direction. In the first half of this review, we survey the literature and the current state of the art focused on enabling optical broadband angular selectivity. The mechanisms we review can be classified into three main categories: (i) microscale geometrical optics, (ii) multilayer birefringent materials, and (iii) Brewster modes in plasmonic systems, photonic crystals, and metamaterials. In the second half, we present two categories of potential applications for broadband angularly selective systems. The first category aims at enhancing the efficiency of solar energy harvesting, through photovoltaic process or solar thermal process. The second category aims at enhancing light extracting efficiency and detection sensitivity. Finally, we discuss the most prominent challenges in broadband angular selectivity and some prospects on how to solve these challenges.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2016

Numerical investigation of shock interaction with one-dimensional transverse array of particles in air

In this paper, we present numerical simulations of shock propagation in air over a one-dimensional transverse array of particles. Simulations are carried out by varying the particle spacing and shock Mach number. We compute the unsteady inviscid drag coefficient as a function of time and make relevant comparisons to that for a single particle. We find that deviations in the drag coefficient in time from that of a single particle can be correlated to the acoustic-particle interaction time. Finally, we investigate and classify the interaction of the bow shocks in front of the transverse array of particles.

42 ENGINEERING↗
DOE OSTI2016

Optical spectroscopy of molecular junctions: Nonequilibrium Green’s functions perspective

We consider optical spectroscopy of molecular junctions from the quantum transport perspective when radiation field is quantized and optical response of the system is simulated as photon flux. Using exact expressions for photon and electronic fluxes derived within the nonequilibrium Green function (NEGF) methodology and utilizing fourth order diagrammatic perturbation theory (PT) in molecular coupling to radiation field, we perform simulations employing realistic parameters. Results of the simulations are compared to the bare PT which is usually employed in studies on nonlinear optical spectroscopy to classify optical processes. We show that the bare PT violates conservation laws, while flux conserving NEGF formulation mixes optical processes.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2021

Measuring the electron temperature and identifying plasma detachment using machine learning and spectroscopy

A machine learning approach has been implemented to measure the electron temperature directly from the emission spectra of a tokamak plasma. This approach utilized a neural network (NN) trained on a dataset of 1865 time slices from operation of the DIII-D tokamak using extreme ultraviolet/vacuum ultraviolet emission spectroscopy matched with high-accuracy divertor Thomson scattering measurements of the electron temperature, T e . This NN is shown to be particularly good at predicting T e at low temperatures (T e < 10 eV) where the NN demonstrated a mean average error of less than 1 eV. Trained to detect plasma detachment in the tokamak divertor, a NN classifier was able to correctly identify detached states (T e < 5 eV) with a 99% accuracy (an F 1 score of 0.96) at an acquisition rate 10× faster than the Thomson scattering measurement. The performance of the model is understood by examining a set of 4800 theoretical spectra generated using collisional radiative modeling that was also used to predict the performance of a low-cost spectrometer viewing nitrogen emission in the visible wavelengths. Furthermore, these results provide a proof-of-principle that low-cost spectrometers leveraged with machine learning can be used to boost the performance of more expensive diagnostics on fusion devices and be used independently as a fast and accurate T e measurement and detachment classifier.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2021

Machine learning methods for probabilistic locked-mode predictors in tokamak plasmas

A rotating tokamak plasma can interact resonantly with the external helical magnetic perturbations, also known as error fields. This can lead to locking and then to disruptions. We leverage machine learning (ML) methods to predict the locking events. We use a coupled third-order nonlinear ordinary differential equation model to represent the interaction of the magnetic perturbation and the plasma rotation with the error field. This model is sufficient to describe qualitatively the locking and unlocking bifurcations. Here, we explore using ML algorithms with the simulation data and experimental data, focusing on the methods that can be used with sparse datasets. These methods lead to the possibility of the avoidance of locking in real-time operations. We describe the operational space in terms of two control parameters: the magnitude of the error field and the rotation frequency associated with the momentum source that maintains the plasma rotation. The outcomes are quan- tified by order parameters that completely characterize the state, whether locked or unlocked. We use unsupervised ML methods to classify locked/unlocked states and note the usefulness of a certain normalization of the order parameters. Three supervised ML classifiers are used in suite to estimate the probability of locking in the region of control parameter space with hysteresis, i.e., the set of control parameters for which both locked and unlocked states can exist. The results show that a neural network gives the best estimate of the locking probability. An analogy of the present locking model with the van der Waals equation of state is also provided.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2021

The superatomic state beyond conventional magic numbers: Ligated metal chalcogenide superatoms

The field of cluster science is drawing increasing attention due to the strong size and composition-dependent properties of clusters and the exciting prospect of clusters serving as the building blocks for materials with tailored properties. However, identifying a unifying central paradigm that provides a framework for classifying and understanding the diverse behaviors is an outstanding challenge. One such central paradigm is the superatom concept that was developed for metallic and ligand-protected metallic clusters. The periodic electronic and geometric closed shells in clusters result in their properties being based on the stability they gain when they achieve closed shells. This stabilization results in the clusters having a well-defined valence allowing them to be classified as superatoms – thus, extending the periodic table to a third dimension. This perspective focuses on extending the superatomic concept to ligated metal-chalcogen clusters that have recently been synthesized in solutions and form assemblies with counterions that have wide-ranging applications. Here we illustrate that the periodic patterns emerge in the electronic structure of ligated metal-chalcogenide clusters. The stabilization gained by the closing of their electronic shells allows for the prediction of their redox properties. Further investigations reveal how the selection of ligands may control the redox properties of the superatoms. These ligated clusters may serve as chemical dopants for two-dimensional semiconductors to control their transport characteristics. Superatomic molecules of multiple metal-chalcogen superatoms allow for the formation of nano pn junctions ideal for directed transport and photon harvesting. As a result, the perspective outlines future developments, including the synthesis of magnetic superatoms.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Demonstration of object location, classification, and characterization by developed deep learning dust ablation trail analysis code package using plasma jets

Based on deep learning, a Dust Ablation Trail Analysis (DATA) code package is developed to detect dust ablation trails in tokamaks, which is intended to analyze a large amount data of tokamak dusts. To validate and benchmark the DATA code package, 2440 plasma jet images are exploited for the training and test of the deep learning DATA code package, since plasma jets resemble the shape and size of dust ablation clouds in tokamaks. After being trained by 1920 plasma jet images, the DATA code package is able to locate 100% plasma jets, classify plasma jets with the accuracy of >99.9%, and output image skeleton information for classified plasma jets. The DATA code package trained by the plasma jet images is also used to analyze the dust ablation trails captured in the Experimental Advanced Superconducting (EAST) tokamak with the satisfactory performance, further verifying its applicability in the fusion dust ablation investigation. Based on its excellent performance presented here, it is demonstrated that our DATA code package is able to automatically identify and analyze dust ablation trails in tokamaks, which can be used for further detailed investigations, such as the three-dimensional reconstruction of dusts and their ablation trails.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗