Development of multivariate function models by least squares
Multivariate functional models for least squares approximation of error tolerance
LEAST SQUARES METHOD↗
SEARCH · Search NASA
Results for “function 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.
Deep probabilistic direction prediction in 3D with applications to directional dark matter detectors
Abstract We present the first method to probabilistically predict 3D direction in a deep neural network model. The probabilistic predictions are modeled as a heteroscedastic von Mises-Fisher distribution on the sphere S 2 , giving a simple way to quantify aleatoric uncertainty. This approach generalizes the cosine distance loss which is a special case of our loss function when the uncertainty is assumed to be uniform across samples. We develop approximations required to make the likelihood function and gradient calculations stable. The method is applied to the task of predicting the 3D directions of electrons, the most complex signal in a class of experimental particle physics detectors designed to demonstrate the particle nature of dark matter and study solar neutrinos. Using simulated Monte Carlo data, the initial direction of recoiling electrons is inferred from their tortuous trajectories, as captured by the 3D detectors. For 40 keV electrons in a 70% He 30% CO 2 gas mixture at STP, the new approach achieves a mean cosine distance of 0.104 (26 ∘ ) compared to 0.556 (64 ∘ ) achieved by a non-machine learning algorithm. We show that the model is well-calibrated and accuracy can be increased further by removing samples with high predicted uncertainty. This advancement in probabilistic 3D directional learning could increase the sensitivity of directional dark matter detectors.
Computer Science↗
Coagulation algorithms with size binning
The Smoluchowski equation describes the time evolution of an aerosol particle size distribution due to aggregation or coagulation. Any algorithm for computerized solution of this equation requires a scheme for describing the continuum of aerosol particle sizes as a discrete set. One standard form of the Smoluchowski equation accomplishes this by restricting the particle sizes to integer multiples of a basic unit particle size (the monomer size). This can be inefficient when particle concentrations over a large range of particle sizes must be calculated. Two algorithms employing a geometric size binning convention are examined: the first assumes that the aerosol particle concentration as a function of size can be considered constant within each size bin; the second approximates the concentration as a linear function of particle size within each size bin. The output of each algorithm is compared to an analytical solution in a special case of the Smoluchowski equation for which an exact solution is known . The range of parameters more appropriate for each algorithm is examined.
Statton, David M.↗
Fully thermal meta-GGA exchange correlation free-energy density functional
The application of density functional theory to materials in the warm dense matter regime has motivated the development of exchange-correlation functionals which incorporate proper, explicit temperature dependence. Previous work has yielded fully-thermal exchange-correlation free energy functionals at the local density approximation (LDA) and generalized gradient approximation (GGA) levels of refinement. Recently an additive thermal correction scheme was utilized to construct a meta-GGA exchange-correlation (XC) functional in which thermal effects are treated at the GGA level. Here, the f TSCAN free-energy XC functional presented here includes thermal effects through the meta-GGA level in the context of the SCAN (strongly constrained and appropriately normed) ground-state functional. The f TSCAN functional provides generality while achieving similar performance to a thermal GGA functional at high temperatures, e.g. pressures within 1% of path integral Monte Carlo simulations of warm dense hydrogen, and a significant improvement over ground-state functionals. At low temperatures, f TSCAN demonstrates improvements in accuracy relative to lower-level and deorbitalized functionals, indicating that calculations using f TSCAN may be expected to perform well across experimentally relevant densities and pressures.
36 MATERIALS SCIENCE↗
Temperature dependence of optical properties of GaAs
The effect of temperature on the optical properties of GaAs was investigated using spectroscopic ellipsometry measurements, between room temperature and about 610 C in increments of 50 C, of pseudodielectric functions and related optical constants of GaAs. A quantitative analysis of the pseudodielectric function spectrum was carried out using a harmonic-oscillator approximation (HOA) to fit the measured dielectric functions. Good fits were obtained with this model, which provides a convenient means of reproducing the GaAs dielectric function at any temperature, by using the temperature-dependent oscillator parameters. The HOA analysis also provides information about band-gap variation with temperature. Using the measured optical constants at a number of fixed temperatures, an algorithm was developed for computing the dielectric function spectrum at an arbitrary temperature in the range 22-610 C.
Yao, Huade↗
Three-dimensional organization of otolith-ocular reflexes in rhesus monkeys. III. Responses To translation
The three-dimensional (3-D) properties of the translational vestibulo-ocular reflexes (translational VORs) during lateral and fore-aft oscillations in complete darkness were studied in rhesus monkeys at frequencies between 0.16 and 25 Hz. In addition, constant velocity off-vertical axis rotations extended the frequency range to 0.02 Hz. During lateral motion, horizontal responses were in phase with linear velocity in the frequency range of 2-10 Hz. At both lower and higher frequencies, phase lags were introduced. Torsional response phase changed more than 180 degrees in the tested frequency range such that torsional eye movements, which could be regarded as compensatory to "an apparent roll tilt" at the lowest frequencies, became anticompensatory at all frequencies above approximately 1 Hz. These results suggest two functionally different frequency bandwidths for the translational VORs. In the low-frequency spectrum (<<0.5 Hz), horizontal responses compensatory to translation are small and high-pass-filtered whereas torsional response sensitivity is relatively frequency independent. At higher frequencies however, both horizontal and torsional response sensitivity and phase exhibit a similar frequency dependence, suggesting a common role during head translation. During up-down motion, vertical responses were in phase with translational velocity at 3-5 Hz but phase leads progressively increased for lower frequencies (>90 degrees at frequencies <0.2 Hz). No consistent dependence on static head orientation was observed for the vertical response components during up-down motion and the horizontal and torsional response components during lateral translation. The frequency response characteristics of the translational VORs were fitted by "periphery/brain stem" functions that related the linear acceleration input, transduced by primary otolith afferents, to the velocity signals providing the input to the velocity-to-position neural integrator and the oculomotor plant. The lowest-order, best-fit periphery/brain stem model that approximated the frequency dependence of the data consisted of a second order transfer function with two alternating poles (at 0.4 and 7.2 Hz) and zeros (at 0.035 and 3.4 Hz). In addition to clearly differentiator dynamics at low frequencies (less than approximately 0.5 Hz), there was no frequency bandwidth where the periphery/brain stem function could be approximated by an integrator, as previously suggested. In this scheme, the oculomotor plant dynamics are assumed to perform the necessary high-frequency integration as required by the reflex. The detailed frequency dependence of the data could only be precisely described by higher order functions with nonminimum phase characteristics that preclude simple filtering of afferent inputs and might be suggestive of distributed spatiotemporal processing of otolith signals in the translational VORs.
Non-NASA Center↗
Electron distribution function in a slightly ionized moving gas.
Lorentz gas approximation to Boltzmann collision operator generalized for heavy molecule nonMaxwellian gas distribution function, obtaining equilibrium distribution applicable to ionized gas
Mintzer, D.↗
Linear Approximation SAR Azimuth Processing Study
A segmented linear approximation of the quadratic phase function that is used to focus the synthetic antenna of a SAR was studied. Ideal focusing, using a quadratic varying phase focusing function during the time radar target histories are gathered, requires a large number of complex multiplications. These can be largely eliminated by using linear approximation techniques. The result is a reduced processor size and chip count relative to ideally focussed processing and a correspondingly increased feasibility for spaceworthy implementation. A preliminary design and sizing for a spaceworthy linear approximation SAR azimuth processor meeting requirements similar to those of the SEASAT-A SAR was developed. The study resulted in a design with approximately 1500 IC's, 1.2 cubic feet of volume, and 350 watts of power for a single look, 4000 range cell azimuth processor with 25 meters resolution.
Lindquist, R. B.↗
Response function of modulated grid Faraday cup plasma instruments
Modulated grid Faraday cup plasma analyzers are a very useful tool for making in situ measurements of space plasmas. One of their great attributes is that their simplicity permits their angular response function to be calculated theoretically. An expression is derived for this response function by computing the trajectories of the charged particles inside the cup. The Voyager plasma science experiment is used as a specific example. Two approximations to the rigorous response function useful for data analysis are discussed. Multisensor analysis of solar wind data indicates that the formulas represent the true cup response function for all angles of incidence with a maximum error of only a few percent.
Barnett, A.↗
Approximate formulation of redistribution in the Ly-alpha, Ly-beta, H-alpha system
Simple approximate formulas are given for the coupled redistribution of Ly-alpha, Ly-beta, and H-alpha, by using well-defined approximations to an essentially exact formulation. These formulas incorporate all the essential physics including Raman scattering, lower state radiative decay, and correlated terms representing emission during a collision which must be retained in order that the emission coefficients are properly behaved in the line wings. Approximate expressions for the appropriate line broadening parameters are collected. Finally, practical expressions for the source functions are given. These are formulated through newly introduced nonimpact redistribution functions, which are shown to be reasonably approximated by existing (ordinary and generalized) redistribution functions.
Cooper, J.↗
A simple parameterization for quality factor as a function of linear energy transfer
The paper presents a simple analytic approximation of the radiation quality factor (Q) as a function of linear energy transfer for use in radiation protection calculations. The paper also presents estimated quality factors in water for protons over a broad range of incident energies. It is shown that the quality factors are less than unity for all proton energies greater than 13 MeV.
Townsend, Lawrence W.↗
The response function of modulated grid Faraday cup plasma instruments
Modulated grid Faraday cup plasma analyzers are a very useful tool for making in situ measurements of space plasmas. One of their great attributes is that their simplicity permits their angular response function to be calculated theoretically. An expression is derived for this response function by computing the trajectories of the charged particles inside the cup. The Voyager Plasma Science (PLS) experiment is used as a specific example. Two approximations to the rigorous response function useful for data analysis are discussed. The theoretical formulas were tested by multi-sensor analysis of solar wind data. The tests indicate that the formulas represent the true cup response function for all angles of incidence with a maximum error of only a few percent.
Barnett, A.↗
Seasonal Variability of Middle Latitude Ozone in the Lowermost Stratosphere Derived from Probability Distribution Functions
We present a study of the distribution of ozone in the lowermost stratosphere with the goal of characterizing the observed variability. The air in the lowermost stratosphere is divided into two population groups based on Ertel's potential vorticity at 300 hPa. High (low) potential vorticity at 300 hPa indicates that the tropopause is low (high), and the identification of these two groups is made to account for the dynamic variability. Conditional probability distribution functions are used to define the statistics of the ozone distribution from both observations and a three-dimensional model simulation using winds from the Goddard Earth Observing System Data Assimilation System for transport. Ozone data sets include ozonesonde observations from northern midlatitude stations (1991-96) and midlatitude observations made by the Halogen Occultation Experiment (HALOE) on the Upper Atmosphere Research Satellite (UARS) (1994- 1998). The conditional probability distribution functions are calculated at a series of potential temperature surfaces spanning the domain from the midlatitude tropopause to surfaces higher than the mean tropical tropopause (approximately 380K). The probability distribution functions are similar for the two data sources, despite differences in horizontal and vertical resolution and spatial and temporal sampling. Comparisons with the model demonstrate that the model maintains a mix of air in the lowermost stratosphere similar to the observations. The model also simulates a realistic annual cycle. Results show that during summer, much of the observed variability is explained by the height of the tropopause. During the winter and spring, when the tropopause fluctuations are larger, less of the variability is explained by tropopause height. This suggests that more mixing occurs during these seasons. During all seasons, there is a transition zone near the tropopause that contains air characteristic of both the troposphere and the stratosphere. The relevance of the results to the assessment of the environmental impact of aircraft effluence is also discussed.
Rood, Richard B.↗
The Unsteady Lift of a Finite Wing
Unsteady lift function for wings of finite aspect ratio have been calculated by approximate methods involving corrections of the aerodynamic inertia and of the angle of the infinite wing. The starting lift of the finite wing is found to be only slightly less than that of the infinite wing; whereas the final lift may be considerably less. The calculations indicate that the distribution of lift near the start is similar to the final distribution. Both the indicia and the oscillating lift functions are given. Approximate operational equivalents of the functions have been devised to facilitate the calculation of lift under various conditions of motion.
Jones, Robert T↗
Small-angle approximation to the transfer of narrow laser beams in anisotropic scattering media
The broadening and the signal power detected of a laser beam traversing an anisotropic scattering medium were examined using the small-angle approximation to the radiative transfer equation in which photons suffering large-angle deflections are neglected. To obtain tractable answers, simple Gaussian and non-Gaussian functions for the scattering phase functions are assumed. Two other approximate approaches employed in the field to further simplify the small-angle approximation solutions are described, and the results obtained by one of them are compared with those obtained using small-angle approximation. An exact method for obtaining the contribution of each higher order scattering to the radiance field is examined but no results are presented.
Box, M. A.↗
Tomography of the rho meson in the QCD instanton vacuum: Transverse momentum dependent parton distribution functions
We analyze the rho meson unpolarized and polarized transverse momentum dependent parton distribution functions (TMDPDFs) in the instanton liquid model (ILM). The corresponding TMDs in ILM are approximated by a constituent quark beam function in the leading Fock state multiplied by a rapidity-dependent factor resulting from the staple-shaped Wilson lines, for fixed longitudinal momentum, transverse separation, and rapidity. At the resolution of the ILM, all of the rho meson TMDs are symmetric in parton 𝑥 for fixed transverse momentum, and Gaussian-like in the transverse momentum for fixed parton 𝑥. The latter is a direct consequence of the profiling of the quark zero modes in the ILM. The evolved TMDs at higher rapidity using the Collins-Soper kernel, and higher resolution using the renormalization group, show substantial skewness towards low parton 𝑥.
Gluons↗
Local thermodynamic mapping for effective liquid density-functional theory
The structural-mapping approximation introduced by Lutsko and Baus (1990) in the generalized effective-liquid approximation is extended to include a local thermodynamic mapping based on a spatially dependent effective density for approximating the solid phase in terms of the uniform liquid. This latter approximation, called the local generalized effective-liquid approximation (LGELA) yields excellent predictions for the free energy of hard-sphere solids and for the conditions of coexistence of a hard-sphere fcc solid with a liquid. Moreover, the predicted free energy remains single valued for calculations with more loosely packed crystalline structures, such as the diamond lattice. The spatial dependence of the weighted density makes the LGELA useful in the study of inhomogeneous solids.
Kyrlidis, Agathagelos↗
Monte Carlo Explicitly Correlated Second-Order Many-Body Green’s Function Calculations of Semiconductor Band Gaps
A systematically converging series of ab initio, post-density-functional, size-consistent, electron-correlated approximations is desired for predictive computing of felectronic band structures of insulating, semiconducting, and metallic solids. A series that meets all of these desiderata (except the applicability to metals) is ab initio many-body Green's function theory based on Gaussian-type-orbital (GTO) basis sets. Here, its leading-order approximation, the second-order Green's function (GF2) method in the diagonal and frequency-independent approximations with the aug-cc-pVDZ basis set, is applied to the fundamental band gaps of three semiconductors (diamond, silicon, and silicon carbide in the zincblende structure) using cluster models. Corrections are made to the basis-set-incompleteness errors by the explicit-correlation (F12) ansatz (GF2-F12) for the valence band edges. The crystals are modeled as surface-passivated clusters of increasing sizes, whose wave functions are expanded by up to 2709 GTO basis functions. Immense computational costs of these calculations are overcome by the highly scalable stochastic algorithm of the Monte Carlo GF2-F12 method, whose operation cost per state increases only as a cubic power of system size, which has a tiny memory footprint and easily achieves near-perfect parallel efficiency on thousands of CPUs or on hundreds of GPUs. The correlated, F12-corrected highest-occupied and lowest-unoccupied molecular-orbital energy (HOMO-LUMO) gap is 5.78 ± 0.07 eV for C 87 H 76 as compared with the experimental value of the fundamental (indirect) band gap of bulk diamond at 5.48 eV. The correlated, F12-corrected HOMO-LUMO gaps for Si 75 H 76 and Si 32 C 43 H 76 are 2.56 ± 0.15 eV and 3.50 ± 0.12 eV, respectively, which are expected to decrease further with increasing cluster sizes. As a result, the experimental fundamental (indirect) band gaps of bulk silicon and silicon carbide are 1.17 eV and 2.42 eV, respectively.
75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗