Search NASA⌕ Search

SEARCH · Search NASA

Results for “Universal approximation”

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 55 records · Page 3
All sourcesNASA NTRSDOE OSTI
NASA NTRS1994

The Local Supercluster as a test of cosmological models

The Local Supercluster kinematic properties (the Local Group infall toward the Virgo Cluster and galaxy density distribution about the Virgo Cluster) in various cosmological models are examined utilizing large-scale N-body (PM) simulations 500(exp 3) cells, 250(exp 3) particles, and box size of 400 h(exp -1) Mpc) and are compared to observations. Five models are investigated: (1) the standard, Cosmic Background Explorer Satellite (COBE)-normalized cold dark matter (CDM) model with omega = 1, h = 0.5, and sigma(sub 8) = 1.05; (2) the standard Hot Dark Matter (HDM) model with omega = 1, h = 0.75, and sigma(sub 8) = 1; (3) the tilted CDM model with omega = 1, h = 0.5, n = 0.7, and sigma(sub 8) = 0.5; (4) a CDM + lambda model with omega = 0.3, lambda = 0.7, h = 2/3, and sigma(sub 8) = 2/3; (5) the PBI model with omega = 0.2, h = 0.8, x = 0.1, m = -0.5, and sigma(sub 8) = 0.9. Comparison of the five models with the presently available observational measurements v(sub LG) = 85 - 305 km/s (with mean of 250 km/s), delta(n(sub g))/(n(sub g)-bar) = 1.40 + or - 0.35) suggests that an open universe with omega approximately 0.5 (with or without lambda) and sigma(sub 8) approximately 0.8 is preferred, with omega = 0.3-1.0 (with or without lambda) and sigma(sub 8) = 0.7-1.0 being the acceptable range. At variance with some previous claims based on either direct N-body or spherical nonlinear approaches, we find that a flat model with sigma(sub 8) approximately 0.7-1.0 seems to be reasonably consistent with observations. However, if one favors the low limit of v(sub LG) = 85 km/s, then an omega approximately 0.2-0.3 universe seems to provide a better fit, and flat (omega = 1) models are ruled out at approximately 95% confidence level. On the other hand, if the high limit of v(sub LG) = 350 km/s is closer to the truth, then it appears that omega approximately 0.7-0.8 is more consistent. This test is insensitive to the shape of the power spectrum, but rather sensitive to the normalization of the perturbation amplitude on the relevant scale (e.g., sigma(sub 8)) and omega. We find that neither linear nor nonlinear relations (with spherical symmetry) are good approximations for the relation between radial peculiar velocity and density perturbation, i.e., nonspherical effects and gravitational tidal field are important. The derived omega using either of the two relations is underestimated. In some cases, this error is as large as a factor of 2-4.

Cen, Renyue↗
NASA NTRS1995

A count probability cookbok: Spurious effects and the scaling model

We study the errors brought by finite volume effects and dilution effects on the practical determination of the count probability distribution function P(sub N)(n,l), which is the probability of having N objects in a cell of volume l cubed for a set of average number density n. Dilution effects are particularly revelant to the so-called sparse sampling strategy. This work is mainly done in the framework of the Bailan & Schaeffer scaling model, which assumes that the Q-body correlation functions obey the scaling relation Xi(sub Q)(lambda r(sub l),....lambda r(sub Q) = lambda(exp -(Q-1)gamma) Xi(sub Q)(r(sub 1),....r(sub Q)). We use three synthetic samples as references to perform our analysis: a fractal generated by a Rayleigh-Levy random walk with approximately 3 x 10(exp 4) objects, a sample dominated by a spherical power-law cluster with approximately 3 x 10(exp 4) objects and a cold dark matter (CDM) universe involving approximately 3 x 10(exp 5) matter particles.

Colombi, S.↗
NASA NTRS1977

Digital overlaying of the universal transverse Mercator grid with LANDSAT data derived products

Picture elements of data from the LANDSAT multispectral scanner are correlated with the universal tranverse Mercator grid. In the procedure, a series of computer modules was used to make approximations of universal transverse Mercator grid locations for all picture elements from the grid locations of a limited number of known control points and to provide display and digital storage of the data. The software has been written in FORTRAN 4 language for a Varian 70-series computer.

Graham, M. H.↗
NASA NTRS1987

Detection of CO(1-0) emission and optical imaging of the Seyfert galaxy/QSO Markarian 231

The detection of CO(J = 1-0) emission and optical imaging of the luminous infrared galaxy Markarian 231 are reported. The galaxy is extremely rich in molecular gas with MT(H2) approximately equal to 1.4 x 10 to the 10th solar masses, approximately 5 times the molecular gas content of the Galaxy. Markarian 231 is the most luminous object in the local universe (z approximately equal to or less than 0.1), with a far-infrared luminosity (lambda = 40-400 microns) of 2.1 x 10 to the 12th solar luminosities. THe CO detection yields a L(FIR)/M(H2) ratio of 150. A deep optical CCD image shows two striking tidal tails with total extent of about 75 kpc. The CCD image strongly suggests that Markarian 231 is an advanced merger system. If the molecular gas is highly concentrated in the nuclear region it may fuel an intense starburst and possibly feed the accretion onto an embedded QSO. The trigger for the intense activity observed in Markarian 231 appears to be the collision of two gas-rich spiral galaxies.

Sanders, D. B.↗
NASA NTRS2002

A Deterministic Interfacial Cyclic Oxidation Spalling Model: Algebraic Approximation, Descriptive Parameters, and Normalized Universal Curve - Part 2

A cyclic oxidation interfacial spalling model has been developed in Part 1. The governing equations have been simplified here by substituting a new algebraic expression for the series (Good-Smialek approximation). This produced a direct relationship between cyclic oxidation weight change and model input parameters. It also allowed for the mathematical derivation of various descriptive parameters as a function of the inputs. It is shown that the maximum in weight change varies directly with the parabolic rate constant and cycle duration and inversely with the spall fraction, all to the 1/2 power. The number of cycles to reach maximum and zero weight change vary inversely with the spall fraction, and the ratio of these cycles is exactly 1:3 for most oxides. By suitably normalizing the weight change and cycle number, it is shown that all cyclic oxidation weight change model curves can be represented by one universal expression for a given oxide scale.

Smialek, James L.↗
DOE OSTI2020

Wormhole calculus, replicas, and entropies

We investigate contributions of spacetime wormholes, describing baby universe emission and absorption, to calculations of entropies and correlation functions, for example those based on the replica method. We find that the rules of the “wormhole calculus”, developed in the 1980s, together with standard quantum mechanical prescriptions for computing entropies and correlators, imply definite rules for limited patterns of connection between replica factors in simple calculations. These results stand in contrast with assumptions that all topologies connecting replicas should be summed over, and call into question the explanation for the latter. In a “free” approximation baby universes introduce probability distributions for coupling constants, and we review and extend arguments that successive experiments in a “parent” universe increasingly precisely fix such couplings, resulting in ultimately pure evolution. Once this has happened, the nontrivial question remains of how topology-changing effects can modify the standard description of black hole information loss.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2025

Parametric matrix models

We present a general class of machine learning algorithms called parametric matrix models. In contrast with most existing machine learning models that imitate the biology of neurons, parametric matrix models use matrix equations that emulate physical systems. Similar to how physics problems are usually solved, parametric matrix models learn the governing equations that lead to the desired outputs. Parametric matrix models can be efficiently trained from empirical data, and the equations may use algebraic, differential, or integral relations. While originally designed for scientific computing, we prove that parametric matrix models are universal function approximators that can be applied to general machine learning problems. After introducing the underlying theory, we apply parametric matrix models to a series of different challenges that show their performance for a wide range of problems. For all the challenges tested here, parametric matrix models produce accurate results within an efficient and interpretable computational framework that allows for input feature extrapolation.

Computational science↗
DOE OSTI2025

Universal control in bosonic systems with weak Kerr nonlinearities

Resonators with weak single-photon self-Kerr nonlinearities can theoretically be used to prepare Fock states in the presence of a loss much larger than their nonlinearities. Two necessary ingredients are large displacements and a two-photon (parametric) drive. Here, in this study, we find that these systems can be controlled to achieve any desired gate operation in a finite-dimensional subspace (whose dimensionality can be chosen at will). Moreover, we show that the two-photon driving requirement can be relaxed and that full controllability is achievable with only one-photon (linear) drives. We make use of both Trotter-Suzuki decompositions and gradient-based optimization to find control pulses for a desired gate, which reduces the computational overhead by using a small blockaded subspace. We also discuss the infidelity arising from input power limitations in realistic settings, as well as from corrections to the rotating-wave approximation. Our universal control protocol opens the possibility for quantum information processing using a wide range of lossy systems with weak nonlinearities.

Yuan, Ming [Univ. of Chicago, IL (United States)] ↗
NASA NTRS2013

Quiescent Galaxies in the 3D-HST Survey: Spectroscopic Confirmation of a Large Number of Galaxies with Relatively Old Stellar Populations at Z approx. 2

Quiescent galaxies at zeta approximately 2 have been identified in large numbers based on rest-frame colors, but only a small number of these galaxies have been spectroscopically confirmed to show that their rest-frame optical spectra show either strong Balmer or metal absorption lines. Here, we median stack the rest-frame optical spectra for 171 photometrically quiescent galaxies at 1.4 less than z less than 2.2 from the 3D-HST grism survey. In addition to H(Beta) (lambda 4861 Angstroms), we unambiguously identify metal absorption lines in the stacked spectrum, including the G band (lambda 4304 Angstroms), Mg I (lambda 5175 Angstroms), and Na i (lambda 5894 Angstroms). This finding demonstrates that galaxies with relatively old stellar populations already existed when the universe was approximately 3 Gyr old, and that rest-frame color selection techniques can efficiently select them. We find an average age of 1.3(+0.1/−0.3) Gyr when fitting a simple stellar population to the entire stack. We confirm our previous result from medium-band photometry that the stellar age varies with the colors of quiescent galaxies: the reddest 80% of galaxies are dominated by metal lines and have a relatively old mean age of 1.6(+0.5/−0.4) Gyr, whereas the bluest (and brightest) galaxies have strong Balmer lines and a spectroscopic age of 0.9(+0.2/−0.1) Gyr. Although the spectrum is dominated by an evolved stellar population, we also find [O III] and Hβ emission. Interestingly, this emission is more centrally concentrated than the continuum with L(sub OIII) = 1.7 +/- 0.3 × 10(exp 40 erg s−1, indicating residual central star formation or nuclear activity.

Quiescent↗
DOE OSTI2025

Direct NeTS sampling of nuclear graphite $S(α, β, T)$ in Serpent

For advanced reactor applications, Neural Thermal Scattering (NeTS) modules were developed to predict the thermal scattering law (TSL or $S(α, β, T)$) of a nuclear graphite neutron moderator. NeTS are multi-layer, feedforward artificial neural networks, which act as universal function approximators designed for TSL datasets. In this case, a 4-layer neural network with 164 neurons per layer is trained using FLASSH evaluated data in PyTorch and serialized as a torchscript dictionary to predict $S(α, β, T)$ on-the-fly. Relative, absolute and maximum percent deviations of NeTS from File 7 data generated using the FLASSH code are on the order of 0.01%, 0.1% and 1%, respectively, with low inference latencies of 0.000172 s per $S(α, β, T)$ at a given temperature. Capturing the full dimensionality of possible inelastic neutron-lattice interactions, NeTS functionality is embedded in the Serpent Monte Carlo code, where $S(α, β, T)_{NeTS}$ sampling is conducted on-the-fly and compared to ACE look-up-tables for predicting TREAT criticality. k-eff differences between sampling algorithms of 6 pcm are observed and are within the order of Monte Carlo uncertainty. Compared to discrete and continuous-energy ACE files (30 MB and 131 MB per temperature), the NeTS format is on the order of 200–300 kB for a continuous-temperature, interpolation-free representation of $S(α, β, T)$ and cross sections. NeTS-in-Serpent runtimes comparable with ACE look-up tables are achieved by scaling NeTS for high performance computing architectures with hybrid OpenMP + MPI parallelization. This work validates a novel, self-contained reactor physics framework for predictive cross sections, and demonstrates a general methodology for embedding modern machine learning libraries within existing neutronic analysis frameworks.

Nuclear Criticality Safety Program (NCSP)↗
DOE OSTI2026

Nonlinear thermodynamic computing out of equilibrium

We present the design for a thermodynamic computer that can perform arbitrary nonlinear calculations in or out of equilibrium. Simple thermodynamic circuits, fluctuating degrees of freedom in contact with a thermal bath and confined by a quartic potential, display an activity that is a nonlinear function of their input. Such circuits can therefore be regarded as thermodynamic neurons, and can serve as the building blocks of networked structures that act as thermodynamic neural networks, universal function approximators whose operation is powered by thermal fluctuations. We simulate a digital model of a thermodynamic neural network, and show that its parameters can be adjusted by genetic algorithm to perform nonlinear calculations at specified observation times, regardless of whether the system has attained thermal equilibrium. This work expands the field of thermodynamic computing beyond the regime of thermal equilibrium, enabling fully nonlinear computations, analogous to those performed by classical neural networks, at specified observation times.

Whitelam, Stephen [Lawrence Berkeley National Labo↗
DOE OSTI2023

Physics informed deep neural network embedded in a chemical transport model for the Amazon rainforest

Secondary organic aerosols (SOA) are fine particles in the atmosphere, which interact with clouds, radiation and affect the Earth’s energy budget. SOA formation involves chemistry in gas phase, aqueous aerosols, and clouds. Simulating these chemical processes involve solving a stiff set of differential equations, which are computationally expensive steps for three-dimensional chemical transport models. Deep neural networks (DNNs) are universal function approximators that could be used to represent the complex nonlinear changes in aerosol physical and chemical processes; however, key challenges such as generalizability to extended time periods, preservation of mass balance, simulating sparse model outputs, and maintaining physical constraints have limited their use in atmospheric chemistry. Here, we develop an approach of using a physics-informed DNN that overcomes previous such challenges and demonstrates its applicability for the chemical formation processes of isoprene epoxydiol SOA (IEPOX-SOA) over the Amazon rainforest. The DNN is trained with data generated by simulating IEPOX-SOA over the entire atmospheric column, using the Weather Research and Forecasting Model coupled with Chemistry (WRF-Chem). The trained DNN is then embedded within WRF-Chem to replace the computationally expensive default solver of IEPOX-SOA formation. The trained DNN predictions generalizes well with the default model simulation of the IEPOX-SOA mass concentrations and its size distribution (20 size bins) over several days of simulations in both dry and wet seasons. The embedded DNN reduces the computational expense of WRF-Chem by a factor of 2. Our approach shows promise in terms of application to other computationally expensive chemistry solvers in climate models.

54 ENVIRONMENTAL SCIENCES↗
NASA NTRS1987

The luminosity function of the brightest galaxies in the IRAS survey

Results from a study of the far infrared properties of the brightest galaxies in the IRAS survey are described. There is a correlation between the infrared luminosity and the infrared to optical luminosity ratio and between the infrared luminosity and the far infrared color temperature in these galaxies. The infrared bright galaxies represent a significant component of extragalactic objects in the local universe, being comparable in space density to the Seyferts, optically identified starburst galaxies, and more numerous than quasars at the same bolometric luminosity. The far infrared luminosity in the local universe is approximately 25% of the starlight output in the same volume.

Soifer, B. T.↗
NASA NTRS1987

An overview of the Hubble Space Telescope command and data management system

The Hubble Space Telescope (HST) will be an orbiting observatory which will extend the capability of large earth-based observatories to view the universe by approximately 350 times in volume and approximately seven times in distance. The HST program, which includes the Orbiting Observatory, the Space Telescope Science Institute, and the Space Telescope Operations Control Center, is supported by the Space Shuttle, the Tracking and Data Relay Satellite System, and the NASA Communications Network. This paper describes the Command and Data Management System for the HST spacecraft, including flexibility, operation modes, and end-to-end data flow paths used to meet stringent synchronization and data quality requirements. The results of system tests indicate that the use of concatenated coding techiques consisting of both block and convolutional codes as a protection against random and burst errors has been successful. Diagrams of the HST data system are included.

Clubb, Jerry↗
NASA NTRS1995

The least-action method, cold dark matter, and omega

Peebles has suggested an interesting technique, called the least-action method, to trace positions of galaxies back in time. This method applied on the Local Group galaxies seems to indicate that we live in an omega approximately = 0.1 universe. We have studied a cold dark matter (CDM) N-body simulation with omega = 0.2 and H = 50 km/s/Mpc and compared trajectories traced back by the least-action method with the ones given by the center of mass of the CDM halos. We show that the agreement between these sets of trajectories is at best qualitative. We also show that the line-of-sight peculiar velocities of halos are underestimated. This discrepancy is due to orphans, i.e., CDM particles which do not end up in halos. We vary the value of omega in the least-action method until the line-of-sight velocities agree with the CDM ones. The best value for this omega underestimates one of the CDM simulations by a factor of 4-5.

Dunn, A. M.↗
NASA NTRS2011

Lyman Break Analogs: Constraints on the Formation of Extreme Starbursts at Low and High Redshift

Lyman Break Analogs (LBAs), characterized by high far-UV luminosities and surface brightnesses as detected by GALEX, are intensely star-forming galaxies in the low-redshift universe (z approximately equal to 0.2), with star formation rates reaching up to 50 times that of the Milky Way. These objects present metallicities, morphologies and other physical properties similar to higher redshift Lyman Break Galaxies (LBGs), motivating the detailed study of LBAs as local laboratories of this high-redshift galaxy population. We present results from our recent integral-field spectroscopy survey of LBAs with Keck/OSIRIS, which shows that these galaxies have the same nebular gas kinematic properties as high-redshift LBGs. We argue that such kinematic studies alone are not an appropriate diagnostic to rule out merger events as the trigger for the observed starburst. Comparison between the kinematic analysis and morphological indices from HST imaging illustrates the difficulties of properly identifying (minor or major) merger events, with no clear correlation between the results using either of the two methods. Artificial redshifting of our data indicates that this problem becomes even worse at high redshift due to surface brightness dimming and resolution loss. Whether mergers could generate the observed kinematic properties is strongly dependent on gas fractions in these galaxies. We present preliminary results of a CARMA survey for LBAs and discuss the implications of the inferred molecular gas masses for formation models.

Goncalves, Thiago S.↗
NASA NTRS2017

The James Webb Space Telescope Mission

The universe is 13.7 billion years old. Hubble can look back in time to observe an epoch during which the universe was approximately 1 billion years old, and has seen some galaxies at even earlier epochs.

Greenhouse, Matt↗
NASA NTRSDate not supplied

Comparison of Multivariate Time Series Prediction Techniques for Emulating Noah-LSM Soil Moisture Outputs

Land surface models are crucial tools for many earth science applications including numerical weather prediction, water resource and crop monitoring, and climatological analysis. Given a set of atmospheric forcings, seasonal data, and static parameters, models like Noah-LSM solve for land surface quantities including skin temperature, sensible heat flux, and soil moisture. While these calculations are theoretically robust, they are often computationally expensive. Since artificial neural networks (ANNs) are universal function approximators, they can learn to emulate the output of a deterministic numerical model given a time series of input forcings, with the learned ANN having substantially shorter execution time. The ANN could efficiently parameterize other models, generate ensembles, and provide first-guess inputs for retrievals. As such, with the goal of developing a model that efficiently mimics the output of Noah-LSM given NLDAS2 forcings on a region covering much of the central US, we examine and compare several neural network architectures for the multi-horizon multivariate time series forecasting problem. Recent literature includes a diverse set of approaches including autoregressive architectures like LSTM and GRU, parametric and non-parametric statistical predictors (ForecastNet and MQRNN), self-attention (LSTM-attention-LSTM), and temporal convovlution (DeepTCN). We implement several of these models for the Noah-LSM prediction task, highlighting the features and challenges for each and providing practical insight on the training process.

Mitchell Dodson↗