Search NASA⌕ Search

SEARCH · Search NASA

Results for “computational statistics”

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 973 records · Page 54

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↗

Quantum-classical hybrid algorithm for the simulation of all-electron correlation

While chemical systems containing hundreds to thousands of electrons remain beyond the reach of quantum devices, hybrid quantum-classical algorithms present a promising pathway toward a quantum advantage. Hybrid algorithms treat the exponentially scaling part of the calculation-the static correlation-on the quantum computer and the non-exponentially scaling part-the dynamic correlation-on the classical computer. While a variety of algorithms have been proposed, the dependence of many methods on the total wave function limits the development of easy-to-use classical post-processing implementations. Here, we present a novel combination of quantum and classical algorithms, which computes the all-electron energy of a strongly correlated molecular system on the classical computer from the 2-electron reduced density matrix (2-RDM) evaluated on the quantum device. Significantly, we circumvent the wave function in the all-electron calculations by using density matrix methods that only require input of the statically correlated 2-RDM. Although the algorithm is completely general, we test it with two classical density matrix methods, the anti-Hermitian contracted Schrödinger equation (ACSE) and multiconfiguration pair-density functional theories, using the recently developed quantum ACSE method for simulating the statically correlated 2-RDM. Furthermore, we obtain experimental accuracy for the relative energies of all three benzyne isomers and thereby demonstrate the ability of the developed algorithm to achieve chemically relevant and accurate results on noisy intermediate-scale quantum devices.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Hierarchical off-diagonal low-rank approximation of Hessians in inverse problems, with application to ice sheet model initialization

Obtaining lightweight and accurate approximations of discretized objective functional Hessians in inverse problems governed by partial differential equations (PDEs) is essential to make both deterministic and Bayesian statistical large-scale inverse problems computationally tractable. The cubic computational complexity of dense linear algebraic tasks, such as Cholesky factorization, that provide a means to sample Gaussian distributions and determine solutions of Newton linear systems is a computational bottleneck at large-scale. These tasks can be reduced to log-linear complexity by utilizing hierarchical off-diagonal low-rank (HODLR) matrix approximations. In this work, we show that a class of Hessians that arise from inverse problems governed by PDEs are well approximated by the HODLR matrix format. In particular, we study inverse problems governed by PDEs that model the instantaneous viscous flow of ice sheets. In these problems, we seek a spatially distributed basal sliding parameter field such that the flow predicted by the ice sheet model is consistent with ice sheet surface velocity observations. Here, we demonstrate the use of HODLR Hessian approximation to efficiently sample the Laplace approximation of the posterior distribution with covariance further approximated by HODLR matrix compression. Computational studies are performed which illustrate ice sheet problem regimes for which the Gauss–Newton data-misfit Hessian is more efficiently approximated by the HODLR matrix format than the low-rank (LR) format. We then demonstrate that HODLR approximations can be favorable, when compared to global LR approximations, for large-scale problems by studying the data-misfit Hessian associated with inverse problems governed by the first-order Stokes flow model on the Humboldt glacier and Greenland ice sheet.

97 MATHEMATICS AND COMPUTING↗

Parity-violating and anisotropic correlations in pseudoscalar inflation

A pseudo-scalar inflaton field can have interesting phenomenological signatures associated with parity violation. The existing analyses of these signatures typically assume statistical isotropy. In the present work we instead investigate the possibility that a pseudo-scalar inflaton is coupled to a vector field carrying a small but non-negligible vacuum expectation value (vev) coherent over our Hubble patch. We show that, in such case, correlators involving the primordial curvature perturbations and gravitational waves violate both statistical isotropy and parity symmetry. We compute the Cosmic Microwave Background (CMB) temperature anisotropies (T) and polarization (E/B) generated by these primordial modes. The CMB two-point correlation functions present distinct signals of broken rotational and parity invariance. Specifically, we find non-vanishing TT, TE, EE and BB correlators between ℓ{sub 1} and ℓ{sub 2} = ℓ{sub 1} ± 1 multipoles, and non-vanishing TB and EB correlators between ℓ{sub 1} and ℓ{sub 2} = ℓ{sub 1} ± 2 multipoles. Such signatures are specific of the models under consideration and they cannot be generated if one of parity and isotropy is preserved. As a specific example we consider the simple case in which the vector field has just an ''electric'' background component decaying in the standard way as a{sup −2}. In this case a strong scale-dependent quadrupolar modulation of the primordial power spectra is generated and we find that almost noiseless data of the large-scale temperature and E-mode polarization anisotropies (like, e.g., the ones provided by WMAP or Planck) should be able to constrain the quadrupolar amplitude coefficients g{sub 2M} of the primordial scalar power spectrum (normalized at the pivot scale comparable to the present horizon size k{sup −1}{sub 0} = 14 Gpc) down to g{sub 2M} = 30 (68%CL)

79 ASTRONOMY AND ASTROPHYSICS↗

Deep neural networks for energy and position reconstruction in EXO-200

Here, we apply deep neural networks (DNN) to data from the EXO-200 experiment. In the studied cases, the DNN is able to reconstruct the relevant parameters—total energy and position—directly from raw digitized waveforms, with minimal exceptions. For the first time, the developed algorithms are evaluated on real detector calibration data. The accuracy of reconstruction either reaches or exceeds what was achieved by the conventional approaches developed by EXO-200 over the course of the experiment. Most existing DNN approaches to event reconstruction and classification in particle physics are trained on Monte Carlo simulated events. Such algorithms are inherently limited by the accuracy of the simulation. We describe a unique approach that, in an experiment such as EXO-200, allows to successfully perform certain reconstruction and analysis tasks by training the network on waveforms from experimental data, either reducing or eliminating the reliance on the Monte Carlo.

97 MATHEMATICS AND COMPUTING↗

Reducing model bias in a deep learning classifier using domain adversarial neural networks in the MINERvA experiment

We present a simulation-based study using deep convolutional neural networks (DCNNs) to identify neutrino interaction vertices in the MINERvA passive targets region, and illustrate the application of domain adversarial neural networks (DANNs) in this context. DANNs are designed to be trained in one domain (simulated data) but tested in a second domain (physics data) and utilize unlabeled data from the second domain so that during training only features which are unable to discriminate between the domains are promoted. MINERvA is a neutrino-nucleus scattering experiment using the NuMI beamline at Fermilab. A-dependent cross sections are an important part of the physics program, and these measurements require vertex finding in complicated events. To illustrate the impact of the DANN we used a modified set of simulation in place of physics data during the training of the DANN and then used the label of the modified simulation during the evaluation of the DANN. In conclusion, we find that deep learning based methods offer significant advantages over our prior track-based reconstruction for the task of vertex finding, and that DANNs are able to improve the performance of deep networks by leveraging available unlabeled data and by mitigating network performance degradation rooted in biases in the physics models used for training.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Geant4Reweight: a framework for evaluating and propagating hadronic interaction uncertainties in Geant4

Geant4Reweight is an open-source C++ framework that allows users to weight tracks produced by the Geant4 particle transport Monte Carlo simulation according to hadron interaction cross section variations and estimate uncertainties in Geant4 interaction models by comparing the simulation’s hadron interaction cross section predictions to data. The ability to weight hadron transport as simulated by Geant4 is crucial to the propagation of systematic uncertainties related to secondary hadronic interactions in current and upcoming neutrino oscillation experiments, including MicroBooNE, NOvA, and DUNE, aswell as hadron test beam experiments such as ProtoDUNE. Here, we provide motivation for weighting hadron tracks in Geant4 in the context of systematic uncertainty propagation, a description of Geant4’s transport simulation technique, and a description of our weighting technique and fitting framework in the momentum range 0–10 GeV/c, which is typical for the hadrons produced by neutrino interactions in these experiments.

97 MATHEMATICS AND COMPUTING↗

Computing the small-scale galaxy power spectrum and bispectrum in configuration space

ABSTRACT We present a new class of estimators for computing small-scale power spectra and bispectra in configuration space via weighted pair and triple counts, with no explicit use of Fourier transforms. Particle counts are truncated at $R_0\sim 100\, h^{-1}\, \mathrm{Mpc}$ via a continuous window function, which has negligible effect on the measured power spectrum multipoles at small scales. This gives a power spectrum algorithm with complexity $\mathcal {O}(NnR_0^3)$ (or $\mathcal {O}(Nn^2R_0^6)$ for the bispectrum), measuring N galaxies with number density n. Our estimators are corrected for the survey geometry and have neither self-count contributions nor discretization artefacts, making them ideal for high-k analysis. Unlike conventional Fourier-transform-based approaches, our algorithm becomes more efficient on small scales (since a smaller R0 may be used), thus we may efficiently estimate spectra across k-space by coupling this method with standard techniques. We demonstrate the utility of the publicly available power spectrum algorithm by applying it to BOSS DR12 simulations to compute the high-k power spectrum and its covariance. In addition, we derive a theoretical rescaled-Gaussian covariance matrix, which incorporates the survey geometry and is found to be in good agreement with that from mocks. Computing configuration- and Fourier-space statistics in the same manner allows us to consider joint analyses, which can place stronger bounds on cosmological parameters; to this end we also discuss the cross-covariance between the two-point correlation function and the small-scale power spectrum.

79 ASTRONOMY AND ASTROPHYSICS↗

Rigorous constraints on three-nucleon forces in chiral effective field theory from fast and accurate calculations of few-body observables

We explore the constraints on the three-nucleon force (3NF) of chiral effective field theory (χ EFT) that are provided by bound-state observables in the A = 3 and A = 4 sectors. Our statistically rigorous analysis incorporates experimental error, computational method uncertainty, and the uncertainty due to truncation of the χ EFT expansion at next-to-next-to-leading order. A consistent solution for the 3 H binding energy, the 4 He binding energy and radius, and the 3 H β-decay rate can only be obtained if χ EFT truncation errors are included in the analysis. Here, the β-decay rate is the only one of these that yields a nondegenerate constraint on the 3NF low-energy constants, which makes it crucial for the parameter estimation. We use eigenvector continuation for fast and accurate emulation of no-core shell model calculations of the few-nucleon observables. This facilitates sampling of the posterior probability distribution, allowing us to also determine the distributions of the parameters that quantify the truncation error. We find a χ EFT expansion parameter of Q = 0.33 ± 0.06 for these observables.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Projection algorithm for state preparation on quantum computers

We present an efficient method to prepare states of a many-body system on quantum hardware, first isolating individual quantum numbers and then using time evolution to isolate the energy. Our method in its simplest form requires only one additional auxiliary qubit. The total time evolved for an accurate solution is proportional to the ratio of the spectrum range of the trial state to the gap to the lowest excited state, a substantial improvement over other projection algorithms, and the accuracy increases exponentially with the time evolved. Isolating the quantum numbers is efficient because of the known eigenvalues and increases the gap thus shortening the propagation time required. The success rate of the algorithm, or the probability of producing the desired state, is a simple function of measurement times and phases and is dominated by the square overlap of the original state to the desired state. We present examples from the nuclear shell model and the Heisenberg model. In conclusion, we compare this algorithm to previous algorithms for short evolution times and discuss potential further improvements.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Ideal gas in a strong gravitational field: Area dependence of entropy

We study the thermodynamic parameters like entropy, energy etc. of a box of gas made up of indistinguishable particles when the box is kept in various static background spacetimes having a horizon. We compute the thermodynamic variables using both statistical mechanics as well as by solving the hydrodynamical equations for the system. When the box is far away from the horizon, the entropy of the gas depends on the volume of the box except for small corrections due to background geometry. As the box is moved closer to the horizon with one (leading) edge of the box at about Planck length (L{sub p}) away from the horizon, the entropy shows an area dependence rather than a volume dependence. More precisely, it depends on a small volume A{sub perpendicular}L{sub p}/2 of the box, up to an order O(L{sub p}/K){sup 2} where A{sub perpendicular} is the transverse area of the box and K is the (proper) longitudinal size of the box related to the distance between leading and trailing edge in the vertical direction (i.e. in the direction of the gravitational field). Thus the contribution to the entropy comes from only a fraction O(L{sub p}/K) of the matter degrees of freedom and the rest are suppressed when the box approaches the horizon. Near the horizon all the thermodynamical quantities behave as though the box of gas has a volume A{sub perpendicular}L{sub p}/2 and is kept in a Minkowski spacetime. These effects are: (i) purely kinematic in their origin and are independent of the spacetime curvature (in the sense that the Rindler approximation of the metric near the horizon can reproduce the results) and (ii) observer dependent. When the equilibrium temperature of the gas is taken to be equal to the horizon temperature, we get the familiar A{sub perpendicular}/L{sub p}{sup 2} dependence in the expression for entropy. All these results hold in a D+1 dimensional spherically symmetric spacetime. The analysis based on methods of statistical mechanics and the one based on thermodynamics applied to the gas treated as a fluid in static geometry, lead to the same results showing the consistency. The implications are discussed.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Soft-sphere simulations of a planar shock interaction with a granular bed

Here we consider the problem of shock propagation through a layer of spherical particles. A point particle force model is used to capture the shock-induced aerodynamic force acting upon the particles. The discrete element method (DEM) code liggghts is used to implement the shock-induced force as well as to capture the collisional forces within the system. A volume-fraction-dependent drag correction is applied using Voronoi tessellation to calculate the volume of fluid around each individual particle. A statistically stationary frame is chosen so that spatial and temporal averaging can be performed to calculate ensemble-averaged macroscopic quantities, such as the granular temperature. A parametric study is carried out by varying the coefficient of restitution for three sets of multiphase shock conditions. A self-similar profile is obtained for the granular temperature that is dependent on the coefficient of restitution. A traveling wave structure is observed in the particle concentration downstream of the shock and this instability arises from the volume-fraction-dependent drag force. The intensity of the traveling wave increases significantly as inelastic collisions are introduced. Downstream of the shock, the variance in Voronoi volume fraction is shown to have a strong dependence upon the coefficient of restitution, indicating clustering of particles induced by collisional dissipation. Statistics of the Voronoi volume are computed upstream and downstream of the shock and compared to theoretical results for randomly distributed hard spheres.

42 ENGINEERING↗

Sparsity-Independent Lyapunov Exponent in the Sachdev-Ye-Kitaev Model

The saturation of a recently proposed universal bound on the Lyapunov exponent has been conjectured to signal the existence of a gravity dual. This saturation occurs in the low-temperature limit of the dense Sachdev-Ye-Kitaev (SYK) model, N Majorana fermions with q body ( q > 2 ) infinite-range interactions. We calculate certain out-of-time-order correlators (OTOCs) for N ≤ 64 fermions for a highly sparse SYK model and find no significant dependence of the Lyapunov exponent on sparsity up to near the percolation limit where the Hamiltonian breaks up into blocks. This provides strong support to the saturation of the Lyapunov exponent in the low-temperature limit of the sparse SYK. A key ingredient to reaching N = 64 is the development of a novel quantum spin model simulation library that implements highly optimized matrix-free Krylov subspace methods on graphical processing units. This leads to a significantly lower simulation time as well as vastly reduced memory usage over previous approaches, while using modest computational resources. Strong sparsity-driven statistical fluctuations require both the use of a much larger number of disorder realizations with respect to the dense limit and a careful finite size scaling analysis. The saturation of the bound in the sparse SYK points to the existence of a gravity analog that would enlarge substantially the number of field theories with this feature. Published by the American Physical Society 2024

Physics↗

Mass of 101 Sn and Bayesian extrapolations to the proton drip line

The favorable energy configurations of nuclei at magic numbers of 𝑁 neutrons and 𝑍 protons are fundamental for understanding the evolution of nuclear structure. The 𝑍 = 50 (tin) isotopic chain is a frontier for such studies, with particular interest at and around the doubly magic 100 Sn isotope, for which the mass is a topic of debate. Precise mass values for neutron-deficient isotopes provide necessary anchor points for mass models to test extrapolations near the proton drip line, where experimental studies remain out of reach. In this work, we report a Penning trap mass measurement of 101 Sn . The determined mass excess of −59889.89⁢(96) keV for 101 Sn represents a factor-of-300 improvement over the current precision and indicates that 101 Sn is less bound than previously thought. Mass predictions from a recently developed Bayesian model combination framework employing statistical machine learning and nuclear masses computed within seven global models based on nuclear density functional theory agree within 1⁢𝜎 with experimental masses from the 48 ≤ 𝑍 ≤ 52 isotopic chains. The framework's resilience to new mass data gave confidence in the extrapolation of tin masses down to 𝑁 = 46. Our calculations suggest that 96 Sn is a two-proton drip line nucleus and predict a mass excess of −58090⁢(800) keV for 100 Sn , showing a preference within 1⁢𝜎 for the mass of 100 Sn derived from the 𝛽-delayed 𝑄 value measured at GSI.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

The Power of Exclusion using Automated Osteometric Sorting: Pair‐Matching,

Abstract This study compares the original pair‐matching osteometric sorting model ( J Forensic Sci 2003 ;48:717) against two new models providing validation and performance testing across three samples. The samples include the Forensic Data Bank, USS Oklahoma , and the osteometric sorting reference used within the Defense POW / MIA Accounting Agency. A computer science solution to generating dynamic statistical models across a commingled assemblage is presented. The issue of normality is investigated showing the relative robustness against non‐normality and a data transformation to control for normality. A case study is provided showing the relative exclusion power of all three models from an active commingled case within the Defense POW / MIA Accounting Agency. In total, 14,357,220 osteometric t ‐tests were conducted. The results indicate that osteometric sorting performs as expected despite reference samples deviating from normality. The two new models outperform the original, and one of those is recommended to supersede the original for future osteometric sorting work.

Lynch, Jeffrey James↗

Statistical error of reactor calculations by the Monte Carlo method

Algorithms for calculating the statistical error with allowance for intergenerational correlations are described. The algorithms are constructed on the basis of statistical analysis of the results of computations by the Monte Carlo method. As a result, simple rules for choosing the parameters of the computational techniques, such as the number of simulated generations necessary for attaining the required accuracy and the number of first skipped generations, are elaborated.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Adapting the M3 Surveillance Metrics for an Unknown Baseline

The original M 3 surveillance metrics assume that the baseline is known. In this article, adapted M 3 metrics are presented when the baseline is not known and estimated by available data. Deciding on how much available data is enough is also discussed.

97 MATHEMATICS AND COMPUTING↗