Search NASA⌕ Search

SEARCH · Search NASA

Results for “tensor methods”

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 37 records · Page 2

Systems and methods for tensor scheduling

A technique for efficient scheduling of operations in a program for parallelized execution thereof using a multi-processor runtime environment having two or more processors includes constraining the type or number of loop optimization transforms that may be explored such that memory and processing capacity available for the scheduling task are not exceeded, while facilitating a tradeoff between memory locality, parallelization, and/or data communication between memory modules of the multi-processor runtime environment.

Meister, Benoit J.↗

Experimental investigation of abnormal transverse flow enhancement of α particles in heavy-ion collisions

The mass dependence of the transverse flow for Z = 1–5 fragments from the collisions of 40 Ar + 27 Al, 40 Ar + 48 Ti, and 40 Ar + 58 Ni at 47 MeV/nucleon is investigated experimentally in this article. The transverse flow values are determined using the in-plane components of the fragment transverse momenta, where three conventional methods, i.e., the kinetic flow tensor method, the transverse momentum analysis method, and the azimuthal correlation method, are applied to reconstruct the reaction plane in an event-by-event basis. It is demonstrated from the comparison of the present experimental mass dependent flow measurements and the model simulations using an improved antisymmetrized molecular dynamics model that the experimentally observed abnormal α transverse flow enhancement is closely related to the reaction plane reconstruction procedure in the flow extraction. We further investigate the physical existence of the abnormal α flow behavior using a two-particle azimuthal correlation method, which allows us to provide the relative flow magnitude information with an identification of fragment charge number without the knowledge of the reaction plane differing from the three conventional methods. It is found that the relative flow magnitudes deduced from the two-particle azimuthal correlation functions with an identification of Z, with the correction for the recoil effect imposed by the momentum conservation, show a monotonically increasing trend as a function of fragment charge number, with no exception of the α flow enhancement. These results, in addition to those from the improved antisymmetrized molecular dynamics model simulations, definitely provide experimental evidences for the inexistence of the abnormal α flow behavior in the heavy-ion collisions at the present incident energy region in nature.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

General-Purpose Bayesian Tensor Learning With Automatic Rank Determination and Uncertainty Quantification

A major challenge in many machine learning tasks is that the model expressive power depends on model size. Low-rank tensor methods are an efficient tool for handling the curse of dimensionality in many large-scale machine learning models. The major challenges in training a tensor learning model include how to process the high-volume data, how to determine the tensor rank automatically, and how to estimate the uncertainty of the results. While existing tensor learning focuses on a specific task, this paper proposes a generic Bayesian framework that can be employed to solve a broad class of tensor learning problems such as tensor completion, tensor regression, and tensorized neural networks. We develop a low-rank tensor prior for automatic rank determination in nonlinear problems. Our method is implemented with both stochastic gradient Hamiltonian Monte Carlo (SGHMC) and Stein Variational Gradient Descent (SVGD). We compare the automatic rank determination and uncertainty quantification of these two solvers. We demonstrate that our proposed method can determine the tensor rank automatically and can quantify the uncertainty of the obtained results. We validate our framework on tensor completion tasks and tensorized neural network training tasks.

Bayesian inference↗

An Efficient Mixed-Mode Representation of Sparse Tensors

The Compressed Sparse Fiber (CSF) representation for sparse tensors is a generalization of the Compressed Sparse Row (CSR) format for sparse matrices. For a tensor with d modes, typical tensor methods such as CANDECOMP/PARAFAC decomposition (CPD) require a sequence of d tensor computations, where efficient memory access with respect to different modes is required for each of them.The straightforward solution is to use d distinct representations of the tensor, with each one being efficient for one of the d computations. However, a d-fold space overhead is often unacceptable in practice, especially with memory-constrained GPUs. In this paper, we present a mixed-mode tensor representation that partitions the tensor’s nonzero elements into disjoint sections, each of which is compressed to create fibers along a different mode. Experimental results on the latest generation of GPU device demonstrate that better performance can be achieved by using the mixed-mode representation, while utilizing only a small fraction of the space required to keep d distinct CSF representations.

Nisa, Israt↗

PASTA: A Parallel Sparse Tensor Algorithm Benchmark Suite

Tensor methods have gained increasingly attention from various applications, including machine learning, quantum chemistry, healthcare analytics, social network analysis, data mining, and signal processing, to name a few. Sparse tensors and their algorithms become critical to further improve the performance of these methods and enhance the interpretability of their output. This work presents a sparse tensor algorithm benchmark suite (PASTA) for single- and multi-core CPUs. To the best of our knowledge, this is the first benchmark suite for sparse tensor world. PASTA targets on: 1) helping application users to evaluate different computer systems using its representative computational workloads; 2) providing insights to better utilize existed computer architecture and systems and inspiration for the future design. This benchmark suite will be publicly released.

HPC, sparse tensor, benchmark↗

Nonnegative tensor decomposition with custom clustering for microphase separation of block copolymers

High–dimensional datasets are becoming ubiquitous in many applications and therefore unsupervised tensor methods to interrogate them are needed. Here, we report a new unsupervised machine learning (ML) approach (NTFk) based on nonnegative tensor factorization integrated with a custom k–means clustering. We demonstrate the ability of NTFk to extracting temporal and spatial features of phase separation of copolymers as they are modeled by self–consistent field theory. Microphase separation of block copolymers has been extensively studied both experimentally and theoretically. However, the interpretation of computer simulations and/or experimental data, representing temporal and spatial changes of molecular species concentration is still a challenging task. Thus, extracting the phase diagram from simulations or experimental data as well as the interpretation of data requires discernment of the model/experimental parameters (such as, temperature, concentrations, the number of molecular species and the interaction between species) impact on the microphase separation process. An attractive and unique aspect of the introduced ML method is that it ensures the nonnegativity of the extracted latent features. Nonnegativity is an essential constraint needed to obtain interpretable and sparse latent features that are parts–based representation of the data. The custom clustering in NTFk serves to estimate the number of latent features in the data.

97 MATHEMATICS AND COMPUTING↗

A space-time tensor formulation for continuum mechanics in general curvilinear, moving, and deforming coordinate systems

Tensor methods are used to express the continuum equations of motion in general curvilinear, moving, and deforming coordinate systems. The space-time tensor formulation is applicable to situations in which, for example, the boundaries move and deform. Placing a coordinate surface on such a boundary simplifies the boundary condition treatment. The space-time tensor formulation is also applicable to coordinate systems with coordinate surfaces defined as surfaces of constant pressure, density, temperature, or any other scalar continuum field function. The vanishing of the function gradient components along the coordinate surfaces may simplify the set of governing equations. In numerical integration of the equations of motion, the freedom of motion of the coordinate surfaces provides a potential for enhanced resolution of the continuum field function. An example problem of an incompressible, inviscid fluid with a top free surface is considered, where the surfaces of constant pressure (including the top free surface) are coordinate surfaces.

Avis, L. M.↗

DNTTD (Distributed Non-Negative Tensor Train Decomposition)

The era of exascale computing opens new venues for innovations and discoveries in many scientific, engineering, and commercial fields. However, with the exa flops also come the extra-large high-dimensional data generated by high performance computing. High-dimensional data is presented as multidimensional arrays, aka tensors. The presence of latent (not directly observable) structures in the tensor allows a unique representation and compression of the data by classical tensor factorization techniques. However, the classical tensor methods are not always stable or they can be exponential in their memory requirements, which makes them not suitable for high-dimensional tensors. Tensor train (TT) is a state-of-the-art tensor network introduced for factorization of high-dimensional tensors. TT transforms the initial high-dimensional tensor in a network of three-dimensional tensors that requires only a linear storage. Many real-world data, such as, density, temperature, population, probability, etc., are non-negative and for an easy interpretation, the algorithms preserving non-negativity are preferred. Here, we introduce a distributed non-negative tensor-train and demonstrate its scalability and the compression on synthetic and real world big datasets.

Bhattarai, Manish↗

Fast multipole method using Cartesian tensor in beam dynamic simulation

Here, the fast multipole method (FMM) using traceless totally symmetric Cartesian tensor to calculate the Coulomb interaction between charged particles will be presented. The Cartesian tensor-based FMM can be generalized to treat other non-oscillating interactions with the help of the differential algebra or the truncated power series algebra. Issues on implementation of the FMM in beam dynamic simulations are also discussed.

43 PARTICLE ACCELERATORS↗

Identifying Isotropic Events Using a Regional Moment Tensor Inversion

In our previous work the deviatoric and isotropic source components for 17 explosions at the Nevada Test Site, as well as 12 earthquakes and 4 collapses in the surrounding region of the western US, were calculated using a regional time-domain full waveform inversion for the complete moment tensor (Dreger et al., 2008; Ford et al., 2008; Ford et al., 2009a). The events separate into specific populations according to their deviation from a pure double-couple and ratio of isotropic to deviatoric energy. The separation allows for anomalous event identification and discrimination between explosions, earthquakes, and collapses. Confidence regions of the model parameters are estimated from the data misfit by assuming normally distributed parameter values. We developed a new Network Sensitivity Solution (NSS) in which the fit of sources distributed over a source-type plot (Hudson et al., 1989) show the resolution of the source parameters. The NSS takes into account the unique station distribution, frequency band, and signal-to-noise ratio of a given event scenario. The NSS compares both a hypothetical pure source (for example an explosion or an earthquake) and the actual data with several thousand sets of synthetic data from a uniform distribution of all possible sources. The comparison with a hypothetical pure source provides the theoretically best-constrained source-type region for a given set of stations, and with it one can determine whether further analysis with the data is warranted. We apply the NSS to a NTS nuclear explosion, and earthquake, as well as the 2006 North Korean explosion, and a nearby earthquake. The results show that explosions and earthquakes are distinguishable, however the solution space depends strongly on the station coverage. Finally, on May 25, 2009 a second North Korean test took place. Our preliminary results show that the explosive nature of the event may be determined using the regional distance moment tensor method. Results indicate that the 2009 event is approximately 5-6 times larger than the earlier test, with an isotropic moment of about 1.8e+22 dyne cm. We perform a series of inversions for pure double-couple, pure explosion, combined double-couple and explosion, full moment tensor, and damped moment tensor inversions to assess the resolution of the isotropic moment of the event.

58 GEOSCIENCES↗

Local magnetic moments in a dinuclear Co{sup 2+} complex as seen by polarized neutron diffraction:Beyond the effective spin-(1/2) model

Polarized neutron diffraction investigations of a paramagnetic molecular dinuclear Co{sup 2+} complex, using the local site susceptibility method, show that the Co{sup 2+} ions carry opposite magnetic moments of 3.1(1) and 3.2(1) {mu}{sub B}, making an angle of 37(1) deg. which is in agreement with the value (39 deg.) provided by the theoretical analysis of the magnetic susceptibility using the model of effective spin 1/2. Polarized neutron diffraction (PND) shows that this dinuclear Co{sup 2+} complex behaves more like a system of two antiferromagnetically coupled ions with spin 3/2, the directions of which are imposed by the distortion axis of the octahedra around each Co{sup 2+} ion due to ligand field. This first application of the local susceptibility tensor method to a molecular compound demonstrates the efficiency of the PND method as a tool for exploring magnetic anisotropy in molecular paramagnets.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Quantum real-time evolution using tensor renormalization group methods

We introduce an approach for approximate real-time evolution of quantum systems using tensor renormalization group (TRG) methods originally developed for imaginary time. We use higher-order TRG to generate a coarse-grained time evolution operator for a 1+1⁢D transverse Ising model with a longitudinal field. We show that the standard tensor norm used for the singular value decomposition-based truncation is degenerate and propose an alternate method to discriminate. We show that it is effective and efficient in evolving Gaussian wave packets for one and two particles in the disordered phase, while ordered phase behavior is more challenging to capture. We compare our algorithm with local simulators for universal quantum computers and discuss possible benchmarking in the near future.

lattice gauge theory↗

Effect of surface-plasmon polaritons on spontaneous emission and intermolecular energy-transfer rates in multilayered geometries

We use a Green's tensor method to investigate the spontaneous emission rate of a molecule and the energy-transfer rate between molecules placed in two types of layered geometries: a slab geometry and a planar waveguide. We focus especially on the role played by surface-plasmon polaritons in modifying the spontaneous emission and energy-transfer rates as compared to free space. In the presence of more than one interface, the surface-plasmon polariton modes split into several branches, and each branch can contribute significantly to modifying the electromagnetic properties of atoms and molecules. Enhancements of several orders of magnitude both in the spontaneous emission rate of a molecule and the energy-transfer rate between molecules are obtained and, by tuning the parameters of the geometry, one has the ability to control the range and magnitude of these enhancements. For the energy-transfer rate interference effects between contributions of different plasmon-polariton branches are observed as oscillations in the distance dependence of this rate.

74 ATOMIC AND MOLECULAR PHYSICS↗

Intermolecular resonance energy transfer in the presence of a dielectric cylinder

Using a Green's tensor method, we investigate the rate of resonance electronic energy transfer between two molecules near a dielectric cylinder. Both the case of a real and frequency-independent dielectric constant {epsilon} and the case of a Drude-Lorentz model for {epsilon}({omega}) are considered, the latter case including dispersion and absorption. If the donor is placed at a fixed position near the cylinder, we find that the energy transfer rate to the acceptor is enhanced compared to its vacuum value in a number of discrete hotspots, centered at the cylinder's surface. In the absence of dispersion and absorption the rate of energy transfer may be enhanced at most a few times. On the other hand, for the Drude-Lorentz model the enhancement may be huge (up to 10{sup 6}) and the hotspots are sharply localized at the surface. We show that these observations can be explained from the fact that in the resonance region of the Drude-Lorentz dielectric surface plasmons occur, which play the dominant role in transferring the electronic energy between the donor and the acceptor. The dependence of the energy transfer rate on the molecular transition frequency is investigated as well. For small intermolecular distances, the cylinder hardly affects the transfer rate, independent of frequency. For larger distances, the frequency dependence is quite strong, in particular in the stop-gap region. The role of the intermolecular distance in the frequency dependence may be explained qualitatively using Heisenberg's uncertainty principle to calculate the spread in the frequencies of the intermediate photons.

74 ATOMIC AND MOLECULAR PHYSICS↗

Amplitude/operator basis in chiral perturbation theory

We establish a systematic construction of the on-shell amplitude/operator basis for Chiral Perturbation Theory (ChPT) in D = 4 spacetime dimensions and with an arbitrary number of flavors N f . For kinematic factors, we employ spinor-helicity variables to construct the soft blocks, which are local amplitudes satisfying the Adler’s zero condition, as well as to take into account the reduction in the kinematic basis due to the Gram determinant, which arises at O(p 10 ) when the number of multiplicity N in an amplitude becomes large: N > D. For flavor factors, we include group-theoretic relations at small N f , N f ≤ N, which decreases the flavor basis. The result is obtained by adapting the Young tensor method of constructing the operator basis for generic effective field theories to the case of non-linearly realized symmetries. Working in the massless quark limit, we present purely mesonic operators for both even- and odd-parity at O(p 6 ) and O(p 8 ) for N = 6 and arbitrary N f , and establish a direct correspondence between the amplitude basis and the operator basis. Furthermore, the redundancy due to the Gram determinant is studied at O(p 10 ) for N = 6, 8 and 10.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Photonic states mixing beyond the plasmon hybridization model

A study is performed on a photonic-state mixing-pattern in an insulator-metal-insulator cylindrical silver nanoshell and its rich variations induced by changes in the geometry and dielectric media of the system, representing the combined influences of plasmon coupling strength and cavity effects. This study is performed in terms of the photonic local density of states (LDOS) calculated using the Green tensor method, in order to elucidate those combined effects. The energy profiles of LDOS inside the dielectric core are shown to exhibit consistently growing number of redshifted photonic states due to an enhanced plasmon coupling induced state mixing arising from decreased shell thickness, increased cavity size effect, and larger symmetry breaking effect induced by increased permittivity difference between the core and the background media. Further, an increase in cavity size leads to increased additional peaks that spread out toward the lower energy regime. A systematic analysis of those variations for a silver nanoshell with a fixed inner radius in vacuum background reveals a certain pattern of those growing number of redshifted states with an analytic expression for the corresponding energy downshifts, signifying a photonic state mixing scheme beyond the commonly adopted plasmon hybridization scheme. Finally, a remarkable correlation is demonstrated between the LDOS energy profiles outside the shell and the corresponding scattering efficiencies.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗