Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “deep”

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 1,675 records · Page 93

Deep quantum circuit simulations of low-energy nuclear states

Numerical simulation is an important method for verifying the quantum circuits used to simulate low-energy nuclear states. However, real-world applications of quantum computing for nuclear theory often generate deep quantum circuits that place demanding memory and processing requirements on conventional simulation methods. Here, we present advances in high-performance numerical simulations of deep quantum circuits to efficiently verify the accuracy of low-energy nuclear physics applications. Our approach employs novel methods for accelerating the numerical simulation including management of simulated mid-circuit measurements to verify projection based state preparation circuits. In this study, we test these methods across a variety of high-performance computing systems and our results show that circuits up to 21 qubits and more than 115,000,000 gates can be efficiently simulated.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Identifying the nature of the QCD transition in relativistic collision of heavy nuclei with deep learning

Using deep convolutional neural network (CNN), the nature of the QCD transition can be identified from the final-state pion spectra from hybrid model simulations of heavy-ion collisions that combines a viscous hydrodynamic model with a hadronic cascade “after-burner”. Two different types of equations of state (EoS) of the medium are used in the hydrodynamic evolution. The resulting spectra in transverse momentum and azimuthal angle are used as the input data to train the neural network to distinguish different EoS. Different scenarios for the input data are studied and compared in a systematic way. A clear hierarchy is observed in the prediction accuracy when using the event-by-event, cascade-coarse-grained and event-fine-averaged spectra as input for the network, which are about 80%, 90% and 99%, respectively. A comparison with the prediction performance by deep neural network (DNN) with only the normalized pion transverse momentum spectra is also made. High-level features of pion spectra captured by a carefully-trained neural network were found to be able to distinguish the nature of the QCD transition even in a simulation scenario which is close to the experiments.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Prospects for beyond the Standard Model physics searches at the Deep Underground Neutrino Experiment: DUNE Collaboration

Abstract The Deep Underground Neutrino Experiment (DUNE) will be a powerful tool for a variety of physics topics. The high-intensity proton beams provide a large neutrino flux, sampled by a near detector system consisting of a combination of capable precision detectors, and by the massive far detector system located deep underground. This configuration sets up DUNE as a machine for discovery, as it enables opportunities not only to perform precision neutrino measurements that may uncover deviations from the present three-flavor mixing paradigm, but also to discover new particles and unveil new interactions and symmetries beyond those predicted in the Standard Model (SM). Of the many potential beyond the Standard Model (BSM) topics DUNE will probe, this paper presents a selection of studies quantifying DUNE’s sensitivities to sterile neutrino mixing, heavy neutral leptons, non-standard interactions, CPT symmetry violation, Lorentz invariance violation, neutrino trident production, dark matter from both beam induced and cosmogenic sources, baryon number violation, and other new physics topics that complement those at high-energy colliders and significantly extend the present reach.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Anomalous electroweak physics unraveled via evidential deep learning

The ever-growing ecosystem of beyond standard model (BSM) calculations and parametrizations has motivated the development of systematic methods for making quantitative cross-comparisons over the wide range of possible models, especially with controllable uncertainties. In this setting, the language of uncertainty quantification (UQ) furnishes useful metrics for assessing statistical overlaps and discrepancies among BSM and related models. In this study, we leverage recent machine learning (ML) developments in evidential deep learning (EDL) for UQ to separate data (aleatoric) and knowledge (epistemic) uncertainties in a model-discrimination setting. We construct several potentially BSM-motivated scenarios for the anomalous electroweak interaction (AEWI) of neutrinos with nucleons in deep inelastic scattering ( v DIS). These scenarios are then quantitatively mapped, as a demonstration, alongside Monte Carlo replicas of the CT18 PDFs used to calculate the $\varDelta \chi ^{2}$ statistic for a typical multi-GeV v DIS experiment, CDHSW. Our framework effectively highlights areas of model agreement and provides a classification of out-of-distribution (OOD) samples. By offering the opportunity to quantitatively understand model overlaps, the approach presented in this work can help facilitate efficient BSM model exploration and exclusion for future New Physics searches.

AI↗

Efficient Training of Deep Neural Operator Networks via Randomized Sampling

Neural operators (NOs) employ deep neural networks to learn the mappings between infinitedimensional function spaces. Deep operator network (DeepONet), a popular NO architecture, has demonstrated success in the real-time prediction of complex dynamics across various scientific and engineering applications. In this work, we introduce a random sampling technique to be adopted during the training of DeepONet, aimed at improving the generalization ability of the model, while significantly reducing the computational time. The proposed approach targets the trunk network of the DeepONet model that outputs the basis functions corresponding to the spatiotemporal locations of the bounded domain on which the physical system is defined. While constructing the loss function, DeepONet training traditionally considers a uniform grid of spatiotemporal points at which all the output functions are evaluated for each iteration. This approach leads to a larger batch size, resulting in poor generalization and increased memory demands, due to the limitations of the stochastic gradient descent (SGD) optimizer. The proposed random sampling over the inputs of the trunk net mitigates these challenges, improving generalization and reducing the memory requirements during training, resulting in significant computational gains. We validate our hypothesis through three benchmark examples, demonstrating substantial reductions in training time while achieving comparable or lower overall test errors relative to the traditional training approach. Our results indicate that incorporating randomization in the trunk network inputs during training enhances the efficiency and robustness of DeepONet, offering a promising avenue for improving the framework’s performance in modeling complex physical systems.

Karumuri, Sharmila [Department of Civil & Systems ↗

Optimizing deep learning hyper-parameters through an evolutionary algorithm. In: MLHPC '15 Proceedings of the Workshop on Machine Learning in High-Performance Computing Environments, Article No. 4

There has been a recent surge of success in utilizing Deep Learning (DL) in imaging and speech applications for its relatively automatic feature generation and, in particular for convolutional neural networks (CNNs), high accuracy classification abilities. While these models learn their parameters through data-driven methods, model selection (as architecture construction) through hyper-parameter choices remains a tedious and highly intuition driven task. To address this, Multi-node Evolutionary Neural Networks for Deep Learning (MENNDL) is proposed as a method for automating network selection on computational clusters through hyper-parameter optimization performed via genetic algorithms.

Young, Steven R.↗

Deep reinforcement learning with online data augmentation to improve sample efficiency for intelligent HVAC control

Deep Reinforcement Learning (DRL) has started showing success in real-world applications such as building energy optimization. Much of the research in this space utilized simulated environments to train RL-agent in an offline mode. Very few research have used DRL-based control in real-world systems due to two main reasons: 1) sample efficiency challenge---DRL approaches need to perform a lot of interactions with the environment to collect sufficient experiences to learn from, which is difficult in real systems, and 2) comfort or safety related constraints---user's comfort must never or at least rarely be violated. In this work, we propose a novel deep Reinforcement Learning framework with online Data Augmentation (RLDA) to address the sample efficiency challenge of real-world RL. We used a time series Generative Adversarial Network (TimeGAN) architecture as a data generator. We further evaluated the proposed RLDA framework using a case study of an intelligent HVAC control. With a ≈28% improvement in the sample efficiency, RLDA framework lays the way towards increased adoption of DRL-based intelligent control in real-world building energy management systems.

Kurte, Kuldeep↗

EPOC Deep Dive Retrospective: A Brief Overview of 7 years of Science Engagement Discussions

Understanding the appropriate ways cyberinfrastructure can be designed, implemented, and executed for scientific use cases requires a deep understanding of the way that researchers and educators interact with technology, and how it may be best implemented to suit their needs. The Engagement and Performance Operations Center (EPOC) has conducted a series of scientific “Deep Dives” of use cases at partner institutions to better understand the requirements for modern scientific innovation across the United States research complex. The results of these activities have revealed gaps in the way that technology has been used to foster research activities. This gap in cyberinfrastructure support has impacts for the overall productivity and innovation possibilities for scientific users.

Zurawski, Jason↗

Altered Regional Homogeneity and Functional Connectivity during Microlesion Period after Deep Brain Stimulation in Parkinson’s Disease

Background. Patients with Parkinson’s disease (PD) undergoing deep brain electrode implantation experience a temporary improvement in motor symptoms before the electrical stimulation begins. We usually call this the microlesion effect (MLE), but the mechanism behind it is not clear. Purpose. This study aimed to assess the alterations in brain functions at the regional and whole-brain levels, using regional homogeneity (ReHo) and functional connectivity (FC), during the postoperative microlesion period after deep brain stimulation (DBS) in PD patients. Method. Resting-state functional MRI data were collected from 27 PD patients before and after the first day of DBS and 12 healthy controls (HCs) in this study. The ReHo in combination with FC analysis was used to investigate the alterations of regional brain activity in all the subjects. Results. There were increased ReHo in the basal ganglia-thalamocortical circuit (left supplementary motor area and bilateral paracentral lobule), whereas decreased ReHo in the default mode network (DMN) (left angular gyrus, bilateral precuneus), prefrontal cortex (bilateral middle frontal gyrus), and the cerebello-thalamocortical (CTC) circuit (Cerebellum_crus2/1_L) after DBS. In addition, we also found abnormal FC in the lingual gyrus, cerebellum, and DMN. Conclusion. Microlesion of the thalamus caused by electrode implantation can alter the activity of the basal ganglia-thalamocortical circuit, prefrontal cortex, DMN, and CTC circuit and induce abnormal FC in the lingual gyrus, cerebellum, prefrontal cortex, and DMN among PD patients. The findings of this study contribute to the understanding of the mechanism of MLE.

Luo, Bei↗

Mechanical Properties of Functionally Graded Concrete Lining for Deep Underground Structures

With the mining depth of coal resources increasing, the thickness of traditional lining for deep mines becomes large. The bearing capacity of the outer lining cannot be fully utilized, so a new type of functionally graded lining structure (FGL) with radial Young’s modulus varying in gradient is proposed. In this study, through theoretical analysis and numerical simulation, the mechanical properties of the functionally graded lining were studied, including the characteristics of the elastic working state and the ultimate bearing state. The influences of the structural parameters of the functionally graded lining, including the inner radius, the thickness, Young’s modulus, the compressive strength of the concrete, Poisson’s ratio, and the number of layers on the mechanical properties were analyzed. The calculation formula of the ultimate bearing capacity of the lining and the calculation formula of the maximum tangential strain at the time of lining failure were put forward. The accuracy of the formula was verified by comparing with the numerical simulation results. The research results provide the basis for the construction of the design theory of a functionally graded lining structure and have great engineering significance for the construction of the kilometer-deep mines in the future.

Li, De-Chun↗

Deep Jointly-Informed neural networks

OJINN Is an easy-to-use deep neural network model that requires fewer user-specified hyper-parameters than traditional neural networks. The algorithm leverages decision trees trained on the data to determine an appropriate deep neural network architectures and weight Initializations. Optional functions also select the learning rate, batch size. and number of training Iterations necessary to create an accurate model.

Humbird, Kelli↗

Deep-Lynx-Python-Package

This software is a python package that interacts with the Application Programming Interface (API) suite provided by Deep Lynx. A python codebase may import this package in order to have access to these methods for communicating with a Deep Lynx instance.

Browning, JerenM↗

Deep Lynx Javascript Package

This software is a JavaScript package that interacts with the Application Programming Interface (API) suite provided by Deep Lynx. A JavaScript or TypeScript codebase may import this package in order to have access to these methods for communicating with a Deep Lynx instance.

Browning, JerenM↗

Deep Learning Denoiser

SAND2022-14744 O The Deep Learning Denoiser uses a trained deep convolutional neural network (CNN) model to decompose an input waveform into a signal and noise waveform. In the approach, the CNN provides a signal mask and a noise mask for an input signal. The Short-Time Fourier Transform (STFT) of the estimated signal is obtained by multiplying the signal mask with the STFT of the input signal. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525.

Koch, Clinton↗

Deep Lynx Matlab Adapter

The Deep Lynx MATLAB Adapter is a Python application that connects the Deep Lynx data warehouse with any MATLAB simulation.

Wilsdon, KatherineN [Idaho National Laboratory] (0↗

deep-taxon v0.1

deep-taxon is a package for training neural networks for classifying genomic sequence data. deep-taxon provides tools for converting genomic sequence data into high performance computing file formats capable of training neural networks at scale.

Tritt, Andrew↗