Search NASA⌕ Search

SEARCH · Search NASA

Results for “TENSOR ANALYSIS”

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 73 records · Page 4
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

New constraints on cosmological modified gravity theories from anisotropic three-point correlation functions of BOSS DR12 galaxies

Here we report a new test of modified gravity theories using the large-scale structure of the Universe. This paper is the first attempt to (1) apply a joint analysis of the anisotropic components of galaxy two- and three-point correlation functions (2 and 3PCFs) to actual galaxy data and (2) constrain the non-linear effects of degenerate higher-order scalar-tensor (DHOST) theories on cosmological scales. Applying this analysis to the Baryon Oscillation Spectroscopic Survey (BOSS) data release 12, we obtain the lower bounds of –1.655 < ξ t and –0.504 < ξ s at the $95{{\ \rm per\ cent}}$ confidence level on the parameters characterizing the time evolution of the tidal and shift terms of the second-order velocity field. These constraints are consistent with GR predictions of ξ t = 15/1144 and ξs = 0. Moreover, they represent a 35-fold and 20-fold improvement, respectively, over the joint analysis with only the isotropic 3PCF. We ensure the validity of our results by investigating various quantities, including theoretical models of the 3PCF, window function corrections, cumulative S/N, Fisher matrices, and statistical scattering effects of mock simulation data. We also find statistically significant discrepancies between the BOSS data and the Patchy mocks for the 3PCF measurement. Finally, we package all of our 3PCF analysis codes under the name hitomi and make them publicly available so that readers can reproduce all the results of this paper and easily apply them to ongoing future galaxy surveys.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2024

Parallel Randomized Tucker Decomposition Algorithms

The Tucker tensor decomposition is a natural extension of the singular value decomposition (SVD) to multiway data. Here, we propose to accelerate Tucker tensor decomposition algorithms by using randomization and parallelization. We present two algorithms that scale to large data and many processors, significantly reduce both computation and communication cost compared to previous deterministic and randomized approaches, and obtain nearly the same approximation errors. The key idea in our algorithms is to perform randomized sketches with Kronecker-structured random matrices, which reduces computation compared to unstructured matrices and can be implemented using a fundamental tensor computational kernel. We provide probabilistic error analysis of our algorithms and implement a new parallel algorithm for the structured randomized sketch. Our experimental results demonstrate that our combination of randomization and parallelization achieves accurate Tucker decompositions much faster than alternative approaches. We observe up to a 16X speedup over the fastest deterministic parallel implementation on 3D simulation data.

Tucker decompositions↗
DOE OSTI2021

Tensor Text-Mining Methods for Malware Identification and Detection, Malware Dynamics Characterization, and Hosts Ranking

Malware is one of the most persistent and costly cyber threats endangering reputation, confidentiality, integrity, and availability for organizations and national security. Consequently, many of the incident detection and prevention systems, and incident responders have begun to utilize machine learning as a helper in the fight against malware and other cyber threats. However, cyber defenders rely on interpretability and generalizability, yet the popular machine learning methods are black-box and often use traditional supervised solutions that do not generalize to novel malware. Therefore, there is a need to improve the existing solutions. At the same time, the majority of the prior research ignored essential evaluation criteria when reporting the results of their methods, which disables the safe reproducibility of the methods in a production environment. Tensor decomposition, on the other hand, enables interpretable unsupervised analysis of the large-scale data for the discovery of hidden patterns. Our findings, performed on real-world and large-scale experiments, show that tensor factorization-based methods yield performance results that surpasses or competes with existing supervised solutions with the added benefit of interpretability and generalizability. With the ability to analyze complex and large-scale data using tensors, we report results that reflect real-world production environments. We propose to develop new game- changing tools for malware identification and characterization that can trace malware evolution, rank the infected or malicious hosts, and streamline the work of incident response teams, malware analysts, and incident detection and prevention systems.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Small- x factorization from effective field theory

We derive a factorization theorem that allows for resummation of small-x logarithms by exploiting Glauber operators in the soft collinear effective field theory. Our analysis is carried out for the hadronic tensor W μν in deep inelastic scattering, and leads to the definition of a new gauge invariant soft function S μν that describes quark and gluon emission in the central region. This soft function provides a new framework for extending resummed calculations for coefficient functions to higher logarithmic orders. Our factorization also defines impact factors by universal collinear functions that are process independent, for instance being identical in small-x DIS and Drell-Yan.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2024

Data-Driven Insights into the Structural Essence of Plasticity in High-Entropy Alloys

The heterogeneous mechanical response of a crystalline alloy with multiple principal elements was investigated using molecular dynamics simulations. The local configuration of the alloy in its quiescent state was characterized by the variables derived from the gyration tensor and the atomic electronegativity. A multivariate analysis identified the geometric and chemical factors that influenced the atomic packing variations. Further, upon straining, the non-affine displacement exhibited spatial heterogeneity. A statistical correlation was established between the local yield events and the specific features of the local configuration. Our findings, validated by the performance metrics analysis, provided a structural criterion for the instability mechanisms in high-entropy alloys (HEAs) and enhanced the understanding of their plasticity.

36 MATERIALS SCIENCE↗
DOE OSTI2021

The effect of elastic strains on the adsorption energy of H, O, and OH in transition metals

The influence of elastic strains on the adsorption of H, O, and OH on the (111) surfaces of 8 fcc (Ni, Cu, Pd, Ag, Pt, Au, Rh, Ir) and on the (0001) surfaces of 3 hcp (Co, Zn, Cd) transition metals was analyzed by means of density functional theory calculations. To this end, surface slabs were subjected to different strain states (uniaxial, biaxial, shear, and a combination of them) up to strains dictated by the mechanical stability limits indicated by phonon calculations. It was found that the adsorption energy followed the predictions of the d-band theory but – surprisingly – the variations in the adsorption energy only depended on the area of the adsorption hole and not on the particular elastic strain tensor applied to achieve this area. The analysis of the electronic structure showed that the applied strains did not modify the shape of the Projected Density of States (PDOS) of the d-orbitals of the transition metals but only led to a shift in the energy levels. Moreover, the presence of the adsorbates on the surfaces led to negligible changes in the PDOS. Thus, the adsorption energies were a function of the Fermi energy which in turn was associated with the change of the area of the adsorption through a general linear law that was valid for all metals. Furthermore, the information in this paper allows the immediate and accurate estimation of the effect of any elastic strain on the adsorption energies of H, O, and OH in 11 transition metals with more than half-filled d-orbitals.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Transversity Distributions and Tensor Charges of the Nucleon: Extraction from Dihadron Production and Their Universal Nature

We perform the first global quantum chromodynamics (QCD) analysis of dihadron production for a comprehensive set of data in electron-positron annihilation, semi-inclusive deep-inelastic scattering, and proton-proton collisions, from which we extract simultaneously the transversity distributions of the nucleon and π + π − dihadron fragmentation functions. We incorporate in our fits known theoretical constraints on transversity, namely, its small- x asymptotic behavior and the Soffer bound. We furthermore show that lattice-QCD results for the tensor charges can be successfully included in the analysis. This resolves the previously reported incompatibility between the tensor charges extracted from dihadron production data and lattice QCD. We also find agreement with results for the transversity and tensor charges obtained from measurements on single-hadron production. Overall, our work demonstrates for the first time the universal nature of all available information for the transversity distributions and the tensor charges of the nucleon. Published by the American Physical Society 2024

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2022

Analytic error analysis of cross section interpolation methods in nodal diffusion codes - I : Theory

This paper discusses two cross section interpolation methods commonly found in popular nodal codes; the partial derivatives and multiple tables models. The motivation for choosing a model, and thus a case matrix structure, is a trade off between accuracy and computational cost. Due to decades of experience, there are default structures that are sufficient for current light water reactor analysis. However, this is not necessarily the case for advanced reactor designs. Therefore, it is advantageous to understand the sources of error in cross section interpolation models so that the quality of a case matrix may be improved. A mathematical framework for these models is presented in this work that provides a more rigorous connection between the nuclear engineering field's cross section interpolation methods and the broader mathematical field of function approximation. The two cross section models examined in this paper were found to utilize Lagrange interpolation and are a subset of Lagrange tensor products. Classical results of Lagrange polynomial error analysis were then applied to the partial derivative and multiple tables models to derive expressions for the total point-wise error. The analytical results classify the total error into two parts: the model form error and interpolation error. Finally, based on our observations, a better foundation for improving the quality of a case matrix is proposed. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2024

Communication Lower Bounds and Optimal Algorithms for Multiple Tensor-Times-Matrix Computation

Multiple tensor-times-matrix (Multi-TTM) is a key computation in algorithms for computing and operating with the Tucker tensor decomposition, which is frequently used in multidimensional data analysis. Here, we establish communication lower bounds that determine how much data movement is required (under mild conditions) to perform the Multi-TTM computation in parallel. The crux of the proof relies on analytically solving a constrained, nonlinear optimization problem. We also present a parallel algorithm to perform this computation that organizes the processors into a logical grid with twice as many modes as the input tensor. We show that, with correct choices of grid dimensions, the communication cost of the algorithm attains the lower bounds and is therefore communication optimal. Finally, we show that our algorithm can significantly reduce communication compared to the straightforward approach of expressing the computation as a sequence of tensor-times-matrix operations when the input and output tensors vary greatly in size.

HBL-inequalities↗
DOE OSTI2020

Generalized Canonical Polyadic Tensor Decomposition

Tensor decomposition is a fundamental unsupervised machine learning method in data science, with applications including network analysis and sensor data processing. This work develops a generalized canonical polyadic (GCP) low-rank tensor decomposition that allows other loss functions besides squared error. For instance, we can use logistic loss or Kullback--Leibler divergence, enabling tensor decomposition for binary or count data. We present a variety of statistically motivated loss functions for various scenarios. We provide a generalized framework for computing gradients and handling missing data that enables the use of standard optimization methods for fitting the model. Furthermore, we demonstrate the flexibility of the GCP decomposition on several real-world examples including interactions in a social network, neural activity in a mouse, and monthly rainfall measurements in India.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

Probing the Electronic Structure of a Thorium Nitride Complex by Solid-State 15N NMR Spectroscopy

The solid-state 15N NMR powder spectra of the thorium nitride complex, [K(18-crown-6)(THF)2][(R2N)3Th(m-15N)(Th(NR2)3] ([K][1-15N], R = SiMe3), and the thorium amide complex, [Th(NR2)3(15NH2)] (2-15N) were recorded. The spectrum for [K][1-15N] represents the first reported solid-state 15N NMR data for an actinide nitride complex. The experimentally measured tensor spans are 847 ppm for [K][1-15N] and 237 ppm for 2-15N. Both shielding tensors exhibit a near-zero asymmetry parameter, which for [K][1-15N] is consistent with a local rotational symmetry of its 15N-labelled nitride ligand. For 2-15N, the lack of asymmetry can be rationalized by a quasi-free Th-NH2 bond rotation in the solid-state. DFT calculations overestimate the tensor span somewhat for [K][1-15N], but provide isotropic shifts in good agreement with both the solid-state and solution values for both complexes. Natural localized molecular orbital (NLMO) analyses of the nuclear shielding reveal that the larger tensor span in [K][1-15N] vs. 2-15N is primarily a consequence of more pronounced covalency of the N-Th bonds, and large spin-orbit coupling due to significant Th 5f orbital contribution to those bonds, impacting the principal components of the shielding tensor perpendicular to the Th-N-Th axis. Overall, our analysis confirms the involvement of the 5f orbitals in Th-N multiple bonds, and further demonstrates the value of solid-state NMR spectroscopy for interrogating actinide-ligand bonding.

actinides, covalent bonding, Nuclear Magnetic Reso↗
DOE OSTI2022

Combining translational and rotational seismic motions to invert local-scale seismic data for time-variable moment tensors: do rotational motions help for high-frequency seismic data produced by underground explosions?

SUMMARY We present an analysis of combining translational and rotational seismic data in an inversion for the time-variable source time functions corresponding to the components of the seismic moment tensor. We conduct a series of numerical experiments where the data are simulated by a combination of an underground explosion and a co-located double couple shear source and recorded on surface-mounted seismometers within 1–2 km of the source. The experiments are designed to mimic explosion seismology experiments, and thus the data are in the 1–10 Hz frequency range and contain very few surface waves. We use a Monte Carlo method to propagate Earth model uncertainty into the estimates of seismic source parameters. In our experiments, we find that the uncertainty of the estimated seismic source parameters increases when we add rotational seismic motions to the inversion when using a constant number of data channels. In this case, the increased degree of uncertainty in the final results is most likely due to the near-surface Earth model uncertainty that we introduce in our simulations. However, for a fixed number of seismic stations, adding rotational seismic motions to the inversion acts to decrease the uncertainty of the estimated seismic source parameters, most likely due to the increase in the number of data channels used in the inversion.

Poppeliers, Christian (ORCID:0000000159526849)↗
DOE OSTI2024

Accelerated Constrained Sparse Tensor Factorization on Massively Parallel Architectures

This study presents the first constrained sparse tensor factorization (cSTF) framework that optimizes and fully offloads computation to massively parallel GPU architectures, and the first performance characterization of cSTF on GPU architectures. In contrast to prior work on tensor factorization, where the matricized tensor times Khatri-Rao product (MTTKRP) is the primary performance bottleneck, our systematic analysis of the cSTF algorithm on GPUs reveals that adding constraints creates an additional bottleneck in the update operation for many real-world sparse tensors. While executing the update operation on the GPU brings significant speedup over its CPU counterpart, it remains a significant bottleneck. To further accelerate the update operation, we propose cuADMM, a new update algorithm that leverages algorithmic and code optimization strategies to minimize both computation and data movement on GPUs. As a result, our framework delivers significantly improved performance compared to prior state-of-the-art. On 10 real-world sparse tensors, our framework achieves geometric mean speedup of 5.1 × (max 41.59 ×) and 7.01 × (max 58.05 ×) on the NIVIDA A100 and H100 GPUs, respectively, over the state-of-the-art SPLATT library running on a 26-core Intel Ice Lake Xeon CPU.

Soh, Yongseok↗
DOE OSTI2022

Classification of (2 + 1) D invertible fermionic topological phases with symmetry

Here we provide a classification of invertible topological phases of interacting fermions with symmetry in two spatial dimensions for general fermionic symmetry groups G f and general values of the chiral central charge c - . Here G f is a central extension of a bosonic symmetry group G b by fermion parity, (-1) F , specified by a second cohomology class [ω 2 ]∈ $\mathscr{H}^2$(Gb,$\mathbb{Z}_2$). Our approach proceeds by gauging fermion parity and classifying the resulting G b symmetry-enriched topological orders while keeping track of certain additional data and constraints. We perform this analysis through two perspectives, using G-crossed braided tensor categories and Spin(2c - ) 1 Chern-Simons theory coupled to a background G gauge field. These results give a way to characterize and classify invertible fermionic topological phases in terms of a concrete set of data and consistency equations, which is more physically transparent and computationally simpler than the more abstract methods using cobordism theory and spectral sequences. Our results also generalize and provide a different approach to the recent classification of fermionic symmetry-protected topological phases by Wang and Gu, which have chiral central charge c - = 0. We show how the tenfold way classification of topological insulators and superconductors fits into our scheme, along with general nonperturbative constraints due to certain choices of c - and G f . Mathematically, our results also suggest an explicit general parametrization of deformation classes of (2 + 1)D invertible topological quantum field theories with G f symmetry.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Updated QCD global analysis of single transverse-spin asymmetries: Extracting $\tilde{H}$, and the role of the Soffer bound and lattice QCD

Here, we present an update to the QCD global analysis of single transverse-spin asymmetries presented in [J. Cammarota et al. (Jefferson Lab Angular Momentum Collaboration), Phys. Rev. D 102, 054002 (2020).] (JAM3D-20). JAM3D-20 simultaneously included transverse momentum dependent and collinear twist-3 observables, both of which are sensitive to quark-gluon-quark correlations in hadrons. In this study we extract for the first time the twist-3 chiral odd fragmentation function $\tilde{H}$ by incorporating the sin φ s modulation data from semi-inclusive deep-inelastic scattering along with its contribution to the single transverse-spin asymmetry in pion production from proton-proton collisions. We also explore the impact of lattice QCD tensor charge calculations and the Soffer bound on our global analysis. We find that both constraints can be accommodated within our results, with $\tilde{H}$ playing a key role in maintaining agreement with the data from proton-proton collisions.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2020

A Sparse Tensor Benchmark Suite for CPUs and GPUs

Tensor computations present significant performance chal- lenges that impact a wide spectrum of applications ranging from machine learning, healthcare analytics, social network analysis, data mining to quantum chemistry and signal processing. Efforts to improve the perfor- mance of tensor computations include exploring data layout, execution scheduling, and parallelism in common tensor kernels. This work presents a benchmark suite for arbitrary-order sparse tensor kernels using state- of-the-art tensor formats: coordinate (COO) and hierarchical coordinate (HiCOO) on CPUs and GPUs. It presents a set of reference tensor kernel implementations that are compatible with real-world tensors and power law tensors extended from synthetic graph generation techniques. We also propose Roofline performance models for these kernels to provide insights of computer platforms from sparse tensor view. This benchmark suite along with the synthetic tensor generator is publicly available.

Li, Jiajia↗
DOE OSTI2025

Investigating Quantum Materials with Half-Polarized Diffraction and magnetic PDF analysis at the HB-2A Neutron Powder Diffractometer

Local magnetic ordering and anisotropy is often central to the emergent behavior and subsequent functional properties in quantum materials and beyond. Neutron powder diffraction provides a straightforward yet extremely powerful technique for quantitative measurements of microscopic magnetic properties. The HB-2A powder diffractometer located at the High Flux Isotope Reactor in ORNL is traditionally utilized for long-range magnetic structure determination. Recently these capabilities have been extended to include methods aimed at accessing local magnetism: Half- polarized neutron powder diffraction (pNPD) and magnetic pair distribution function (mPDF) analysis. These two distinct techniques are possible on HB-2A due to the versatility of the instrument’s reciprocal space coverage, resolution and novel ultra-low temperature multi-sample changers that operate down to dilution refrigerator temperatures. This provides unique capabilities not found on any powder diffraction instrument and is particularly well suited to investigations of magnetic quantum materials. The development and implementation of these techniques will be discussed with a series of science case examples ranging from geometric frustrated magnets to magnetic metal-organic frameworks. Data reduction and analysis tools will be presented that enable the extraction of the local site susceptibility tensor and local spin-spin correlations in real space. Finally, potential combinations of these techniques in the form of half-polarized magnetic pair distribution function (pmPDF) analysis will be considered. Looking forward, HB-2A is undergoing a detector upgrade that will be in the user program by 2026. This will offer an order of magnitude increase in count rates to further aid the development of these often low signal measurements and provide new scientific capabilities.

Neutron Scattering↗
DOE OSTI2024

The Average Spectrum Norm and Near-Optimal Tensor Completion

We propose the average spectrum norm to study the minimum number of measurements required to approximate a multidimensional array (i.e., sample complexity) via low-rank tensor recovery. Our focus is on the tensor completion problem, where the aim is to estimate a multiway array using a subset of tensor entries corrupted by noise. Our average spectrum norm-based analysis provides near-optimal sample complexities, exhibiting dependence on the ambient dimensions and rank that do not suffer from exponential scaling as the order increases.

97 MATHEMATICS AND COMPUTING↗