Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Vectorization”

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 1,027 records · Page 57
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Situation awareness and dynamic ensemble forecasting of abnormal behavior in cyber-physical system

A plurality of monitoring nodes may each generate a time-series of current monitoring node values representing current operation of a cyber-physical system. A feature-based forecasting framework may receive the time-series of and generate a set of current feature vectors using feature discovery techniques. The feature behavior for each monitoring node may be characterized in the form of decision boundaries that separate normal and abnormal space based on operating data of the system. A set of ensemble state-space models may be constructed to represent feature evolution in the time-domain, wherein the forecasted outputs from the set of ensemble state-space models comprise anticipated time evolution of features. The framework may then obtain an overall features forecast through dynamic ensemble averaging and compare the overall features forecast to a threshold to generate an estimate associated with at least one feature vector crossing an associated decision boundary.

Abbaszadeh, Masoud↗
DOE OSTI2020

Extracting dependencies between network assets using deep learning

A network analysis tool receives network flow information and uses deep learning—machine learning that models high-level abstractions in the network flow information—to identify dependencies between network assets. Based on the identified dependencies, the network analysis tool can discover functional relationships between network assets. For example, a network analysis tool receives network flow information, identifies dependencies between multiple network assets based on evaluation of the network flow information, and outputs results of the identification of the dependencies. When evaluating the network flow information, the network analysis tool can pre-process the network flow information to produce input vectors, use deep learning to extract patterns in the input vectors, and then determine dependencies based on the extracted patterns. The network analysis tool can repeat this process so as to update an assessment of the dependencies between network assets on a near real-time basis.

Carroll, Thomas E.↗
DOE OSTI2021

Design Space Exploration of Emerging Memory Technologies for Machine Learning Applications

Memory design space exploration methods study memory systems’ performances and limitations before implementation. The computer memory design space has grown exponentially because of the enormous growth of memory types, memory controllers, and application software. Computer simulators are commonly used for memory design space exploration. However, complex memory simulations take an enormous amount of time. Hence, in this paper, we proposed a machine learning-based design space exploration method for dynamic random-access memory and non-volatile memory systems. We applied our method to the CosmoGAN and LeNet applications to predict the following six memory response parameters: (i) bandwidth, (ii) power, (iii) average latency, (iv) average total latency, (v) memory reads, and (vi) memory writes. Our experimental results show that machine learning models can predict memory response parameter values faster than simulations. We used support vector machine, random forest, and gradient boosting machine learning models. We observed that the support vector machine provides better performance for bandwidth, average latency, and average total latency. The random forest model works better for memory reads and writes. The gradient boosting model provides superior prediction performance for power. We provide a detailed discussion on learning curve characteristics, error analysis, and memory type recommendation.

Hasan, S M Shamimul↗
DOE OSTI2021

Scalable In Situ Computation of Lagrangian Representations via Local Flow Maps

In situ computation of Lagrangian flow maps to enable post hoc time-varying vector field analysis has recently become an active area of research. However, the current literature is largely limited to theoretical settings and lacks a solution to address scalability of the technique in distributed memory. To improve scalability, we propose and evaluate the benefits and limitations of a simple, yet novel, performance optimization. Our proposed optimization is a communication-free model resulting in local Lagrangian flow maps, requiring no message passing or synchronization between processes, intrinsically improving scalability, and thereby reducing overall execution time and alleviating the encumbrance placed on simulation codes from communication overheads. To evaluate our approach, we computed Lagrangian flow maps for four time-varying simulation vector fields and investigated how execution time and reconstruction accuracy are impacted by the number of GPUs per compute node, the total number of compute nodes, particles per rank, and storage intervals. Our study consisted of experiments computing Lagrangian flow maps with up to 67M particle trajectories over 500 cycles and used as many as 2048 GPUs across 512 compute nodes. In all, our study contributes an evaluation of a communication-free model as well as a scalability study of computing distributed Lagrangian flow maps at scale using in situ infrastructure on a modern supercomputer.

Sane, Sudhanshu↗
DOE OSTI2021

Nonlinear Matrix Approximation with Radial Basis Function Components

We introduce and investigate matrix approximation by decomposition into a sum of radial basis function (RBF) components. An RBF component is a generalization of the outer product between a pair of vectors, where an RBF function replaces the scalar multiplication between individual vector elements. Even though the RBF functions are positive definite, the summation across components is not restricted to convex combinations and allows us to compute the decomposition for any real matrix that is not necessarily symmetric or positive definite. We formulate the problem of seeking such a decomposition as an optimization problem with a nonlinear and non-convex loss function. Several modern versions of the gradient descent method, including their scalable stochastic counterparts, are used to solve this problem. We provide extensive empirical evidence of the effectiveness of the RBF decomposition and that of the gradient-based fitting algorithm. While being conceptually motivated by singular value decomposition (SVD), our proposed nonlinear counterpart outperforms SVD by drastically reducing the memory required to approximate a data matrix with the same L2 error for a wide range of matrix types. For example, it leads to 2 to 6 times memory save for Gaussian noise, graph adjacency matrices, and kernel matrices. Moreover, this proximity-based decomposition can offer additional interpretability in applications that involve, e.g., capturing the inner low-dimensional structure of the data, retaining graph connectivity structure, and preserving the acutance of images.

Rebrova, Elizaveta↗
DOE OSTI2021

Mimicking of corruption in images

Methods and systems are provided to generate an uncorrupted version of an image given an observed image that is a corrupted version of the image. In some embodiments, a corruption mimicking (“CM”) system iteratively trains a corruption mimicking network (“CMN”) to generate corrupted images given modeled images, updates latent vectors based on differences between the corrupted images and observed images, and applies a generator to the latent vectors to generate modeled images. The training, updating, and applying are performed until modeled images that are input to the CMN result in corrupted images that approximate the observed images. Because the CMN is trained to mimic the corruption of the observed images, the final modeled images represented the uncorrupted version of the observed images.

Anirudh, Rushil↗
DOE OSTI2021

Search for Non-resonant Higgs Boson Pair Production in the Four Bottom Quark Decay Channel With the CMS Experiment

This dissertation presents a search for non-resonant Higgs boson pair production, focusing on the four bottom quark decay channel. It explores the gluon fusion and vector boson fusion production mechanisms. The analysis is performed with a dataset of proton-proton collisions at a center-of-mass energy of 13 TeV, collected by the CMS detector at the LHC, corresponding to an integrated luminosity of around 138 inverse femtobarns. Innovative techniques in the areas of object identification, event categorization, signal identification, and background modeling are used to maximize the analysis sensitivity. No excess of signal events are observed relative to the background-only expectation, and 95% CL upper limits on the production cross section are set. The observed upper limit on the standard model production cross section is set at 3.6 times the theoretical expectation. At the time of writing this dissertation, it is the most stringent constraint at the LHC from an individual channel. Furthermore, the observed constraint on the coupling modifier of the Higgs boson self-interaction is set between -2.3 and 9.4. The observed constraint on the coupling modifier of the di-vector-boson-di-Higgs-boson interaction is set between -0.1 and 2.2.

Guerrero Ibarra, Guerrero Fernando↗
DOE OSTI2022

Euler-Rodrigues Parameters: A Quantum Circuit to Calculate Rigid-Body Rotations

The use of vectorial parameterization to create geometrical representations in computational models has a large number of applications. One particular application is the calculation of the 3D rotational motion of rigid bodies, that could be used for the spatial location estimation from objects. Provided the algebraic nature of this problem, it could benefit from Quantum Computing, in particular several vectors could be superposed to be transformed with a single operation, providing a quantum processing advantage. In this article, we propose an implementation of a Quantum Computing algorithm to compute Euler-Rodrigues Parameters to model rigid body rotations to transform arbitrary functions, rotating multiple vectors in superposition. We developed this algorithm using Qiskit, taking into account the limitations imposed by the current Noisy Intermediate Scale Quantum (NISQ) devices, such as the reduced number of qubits available and the limited coherence time.

Pelaez, Emilio↗
DOE OSTI2022

Calibration method for a spectral computerized tomography system

A calibration method for an x-ray computerized tomography system and a method of tomographic reconstruction are provided. The calibration method includes steps of measuring at least one point spread function (PSF) at each of a plurality of points, compressing each PSF, and in one or more storing operations, storing the compressed PSFs in a computer-accessible storage medium. The PSF measurements are made in a grid of calibration points in a field of view (FOV) of the system. In the measuring step, an absorber is positioned at each of the calibration points, and an x-ray projection is taken at least once at each of those absorber positions. In the method of tomographic image reconstruction, projection data from an x-ray tomographic projection system are input to an iterative image reconstruction algorithm. The algorithm retrieves and utilizes a priori system information (APSI) The APSI comprises comprising point spread functions (PSFs) of all voxels in a voxelization of the field of view that are compressed in the form of vectors of parameters. For utilization, each retrieved vector of parameters is decompressed so as to generate a discretized PSF.

Jimenez, Jr., Edward Steven↗
DOE OSTI2022

Uncertainty-refined image segmentation under domain shift

A method for digital image segmentation is provided. The method comprises training a neural network for image segmentation with a labeled training dataset from a first domain, wherein a subset of nodes in the neural net are dropped out during training. The neural network receives image data from a second, different domain. A vector of N values that sum to 1 is calculated for each image element, wherein each value represents an image segmentation class. A label is assigned to each image element according to the class with the highest value in the vector. Multiple inferences are performed with active dropout layers for each image element, and an uncertainty value is generated for each image element. The label of any image element with an uncertainty value above a predefined threshold is replaced with a new label corresponding to the class with the next highest value.

Martinez, Carianne↗
DOE OSTI2022

Rare Higgs Processes at CMS and Precision Timing Detector Studies for HL-LHC CMS Upgrade

This thesis describes the search for two rare Higgs processes. The first analysis describes the CMS Run 2 search for $H$ $\rightarrow$ $\mu$$\mu$ decays, with 137.3 fb$^{-1}$ of data at $\sqrt{s}$ = 13 TeV. The analysis targeted four different Higgs production modes: the gluon fusion (ggH), the vector boson fusion (VBF), the Higgs-strahlung process (VH), and the production in association with a pair of top quarks (ttH). Each category used a dedicated machine learning based classifier to separate the signal from the background processes. A combined fit from all these categories saw a slight excess in the data corresponding to 3.0 standard deviations at $M$$_{H}$ = 125.38 GeV, and gave the first evidence for the Higgs boson decay to second-generation fermions. The best-fit signal strength and the corresponding 68% CL interval was found to be +0.17?????? = 1.19 $_{-0.39}^{+0.41}$ (stat)$_{-0.16}^{+0.17}$(syst) at $M$$_{H}$ = 125.38 GeV. The second analysis describes the CMS Run 2 search for 𝐻𝐻 → 𝑏𝑏𝑏𝑏 with highly boosted Higgs bosons. This analysis used a dedicated jet identification algorithm based on graph neural networks (ParticleNet) to identify boosted H→ bb jets. This search targeted the gluon fusion and the vector boson fusion HH production modes, and put constraints on the allowed values of the various Higgs couplings as: 𝜅𝜆 ∈ [−9.9, 16.9] when 𝜅𝑉 = 1, 𝜅2𝑉 = 1; 𝜅𝑉 ∈ [−1.17, −0.79] ∪ [0.81, 1.18] when 𝜅𝜆 = 1, 𝜅2𝑉 = 1; 𝜅2𝑉 ∈ [0.62, 1.41] when 𝜅𝜆 = 1, 𝜅𝑉 = 1. A scenario with 𝜅2𝑉 = 0 was excluded with a significance of 6.3 standard deviations for the first time, when other H couplings are fixed to their SM values. The combined observed (expected) 95% upper limit on the HH production cross section was found to be 9.9 (5.1) × SM. Finally, this thesis also discusses the planned MIP Timing Detector (MTD) upgrade for CMS at the HL-LHC. The MTD will be a time-of-flight (TOF) detector, designed to provide a precision timing information for charged particles using SiPMs + LYSO scintillating crystals, with a time resolution of ∼30 ps. This thesis describes several R&D tests that have been performed for characterizing the sensor properties (time resolution, light yield, etc.) and optimizing the sensor design geometry. This thesis also contains a description of mock test setups for cooling the sensors, since it is known to be an effective way of mitigating the increased dark current rates in the sensors due to radiation damage.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2023

3D segmentation using space carving and 2D convolutional neural networks

A system for generating a 3D segmentation of a target volume is provided. The system accesses views of an X-ray scan of a target volume. The system applies a 2D CNN to each view to generate a 2D multi-channel feature vector for each view. The system applies a space carver to generate a 3D channel volume for each channel based on the 2D multi-channel feature vectors. The system then applies a linear combining technique to the 3D channel volumes to generate a 3D multi-label map that represents a 3D segmentation of the target volume.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Adaptive cache management based on programming model information

A processing system selectively allocates space to store a group of one or more cache lines at a cache level of a cache hierarchy having a plurality of cache levels based on memory access patterns of a software application executing at the processing system. The processing system generates bit vectors indicating which cache levels are to allocate space to store groups of one or more cache lines based on the memory access patterns, which are derived from data granularity and movement information. Based on the bit vectors, the processing system provides hints to the cache hierarchy indicating the lowest cache level that can exploit the reuse potential for a particular data.

Na, Weon Taek↗
DOE OSTI2023

Search for Resonant WZ Production in the Fully Leptonic Final State at 13 TeV With the CMS Detector

A search for an electrically charged, narrow-width diboson (WZ) resonance in the fully leptonic final state using 137:65 fb?1 of data collected by the CMS experiment during the 2016-2018 run period, at the LHC and a center of mass energy of p s = 13 TeV is presented. The search focuses on a channel where the resonance decays to a pair of standard model vector bosons: W and Z. The Z boson subsequently decays to a pair of leptons, either electrons (e) or muons (?). While the W boson decays to a lepton and a neutrino (?). As ? escapes undetected, the neutrino is reconstructed from the missing transverse momentum (Pmisst ) of the event. The analysis is based on a bump-hunt search strategy with background estimation based on stochastically generated simulation events. No significant excess of events is observed above the expected background from the Standard Model and limits are set on the production cross-section of the diboson resonance as a function of the resonance mass at 95% confidence level. Results are interpreted in the context of the Heavy Vector Triplet (HVT) framework.

Hernandez, Andres Vargas↗
DOE OSTI2023

Virtual sensor supervised learning for cyber-attack neutralization

An industrial asset may have monitoring nodes that generate current monitoring node values. A dynamic, resilient estimator may split a temporal monitoring node space into normal and one or more abnormal subspaces associated with different kinds of attack vectors. According to some embodiments, a neutralization model is constructed and trained for each attack vector using supervised learning and the associated abnormal subspace. In other embodiments, a single model is created using out-of-range values for abnormal monitoring nodes. Responsive to an indication of a particular abnormal monitoring node or nodes, the system may automatically invoke the appropriate neutralization model to determine estimated values of the particular abnormal monitoring node or nodes (e.g., by selecting the correct model or using out-of-range values). The series of current monitoring node values from the abnormal monitoring node or nodes may then be replaced with the estimated values.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Finding a contaminant source in a volume of flowing fluid

A method includes: mapping a boundary of a volume of flowing fluid; partitioning the volume by a computational mesh; finding a contaminant location at a first sensor that is disposed within the volume; obtaining a measured velocity of the flowing fluid within the volume; generating a reversed velocity vector field within the mesh, in response to the measured velocity; time stepping the contaminant location from the first sensor along the reversed velocity vector field until the contaminant location intersects the boundary of the volume; and finding a contaminant source at the intersection of the time stepped contaminant location with the boundary of the volume.

Lopez-Marrero, Vanessa↗
DOE OSTI2023

Modified biosensors and biocatalysts and methods of use

Biosensors including a nucleic acid encoding a PcaU protein, a PobR protein, a CatM protein, a PcaR protein, or a TphR protein are provided. In some examples, the biosensors include a promoter regulated by the sensed molecule operably linked to a reporter gene. The biosensors may be included in a vector or in cells including one or more of the biosensors or vectors. Modified chorismate pyruvate lyase (UbiC) and modified paraoxonase (PON1) proteins including one or more amino acid substitutions are provided. Finally, methods of selecting biocatalysts with increased activity including transforming a library of cells expressing a biosensor with one or more nucleic acids encoding one or more mutations in a gene involved in a biosynthesis pathway, determining activity of the reporter protein; and selecting a cell with increased reporter protein activity as expressing a biocatalyst with increased activity are provided.

Jha, Ramesh K.↗
DOE OSTI2023

Generalizing mkFit and its Application to HL-LHC

mkFit is an implementation of the Kalman filter-based track reconstruction algorithm that exploits both thread- and data-level parallelism. In the past few years the project transitioned from the R&D phase to deployment in the Run-3 offline workflow of the CMS experiment. The CMS tracking performs a series of iterations, targeting reconstruction of tracks of increasing difficulty after removing hits associated to tracks found in previous iterations. mkFit has been adopted for several of the tracking iterations, which contribute to the majority of reconstructed tracks. When tested in the standard conditions for production jobs, speedups in track pattern recognition are on average of the order of 3.5x for the iterations where it is used (3-7x depending on the iteration). Multiple factors contribute to the observed speedups, including vectorization and a lightweight geometry description, as well as improved memory management and single precision. Efficient vectorization is achieved with both the icc and the gcc (default in CMSSW) compilers and relies on a dedicated library for small matrix operations, Matriplex, which has recently been released in a public repository. While the mkFit geometry description already featured levels of abstraction from the actual Phase-1 CMS tracker, several components of the implementations were still tied to that specific geometry. We have further generalized the geometry description and the configuration of the run-time parameters, in order to enable support for the Phase-2 upgraded tracker geometry for the HL-LHC and potentially other detector configurations. The implementation strategy and high-level code changes required for the HL-LHC geometry are presented. Speedups in track building from mkFit imply that track fitting becomes a comparably time consuming step of the tracking chain.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗