Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “accuracy”

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 613 records · Page 34
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Numerical investigation of inertial sphere impaction on a cylinder in subcritical flow

We investigated the accuracy of a Lagrangian scheme for simulating both the motion and wall impaction of neutrally buoyant inertial spheres (8.85 ≤ St ≤ 18.04) moving in the subcritical regime (Re = 10,972 and 22,366) of flows past a circular cylinder. The accuracy of the eddy-resolving simulations of the flow field was verified based on available data at nearly comparable flow conditions. The accuracy of the eddy-resolving simulations of the flow field was verified based on available data at nearly comparable flow conditions. The accuracy of the particle tracking scheme was verified with respect to corresponding experimental data in a laboratory flume. Particle dispersion en route to the circular cylinder is mostly driven by the flow disturbance from the particle injector tube. The trends in collision rates are largely explained by the rate of particle entertainment into the boundary layer around the cylinder and the inertial condition of the spheres in this proximity region—both of which can be better assessed by post-processing of the numerical outcomes. Here, the primary incentive to undertake this work was the need for validating a particle-based method for quantifying the hydraulic stressors leading to bodily damage of fish passing flows influenced by hydraulic infrastructure.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2024

Deep energy-pressure regression for a thermodynamically consistent EOS model

Abstract In this paper, we aim to explore novel machine learning (ML) techniques to facilitate and accelerate the construction of universal equation-Of-State (EOS) models with a high accuracy while ensuring important thermodynamic consistency. When applying ML to fit a universal EOS model, there are two key requirements: (1) a high prediction accuracy to ensure precise estimation of relevant physics properties and (2) physical interpretability to support important physics-related downstream applications. We first identify a set of fundamental challenges from the accuracy perspective, including an extremely wide range of input/output space and highly sparse training data. We demonstrate that while a neural network (NN) model may fit the EOS data well, the black-box nature makes it difficult to provide physically interpretable results, leading to weak accountability of prediction results outside the training range and lack of guarantee to meet important thermodynamic consistency constraints. To this end, we propose a principled deep regression model that can be trained following a meta-learning style to predict the desired quantities with a high accuracy using scarce training data. We further introduce a uniquely designed kernel-based regularizer for accurate uncertainty quantification. An ensemble technique is leveraged to battle model overfitting with improved prediction stability. Auto-differentiation is conducted to verify that necessary thermodynamic consistency conditions are maintained. Our evaluation results show an excellent fit of the EOS table and the predicted values are ready to use for important physics-related tasks.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

Prediction of histone post-translational modifications using deep learning

Abstract Motivation Histone post-translational modifications (PTMs) are involved in a variety of essential regulatory processes in the cell, including transcription control. Recent studies have shown that histone PTMs can be accurately predicted from the knowledge of transcription factor binding or DNase hypersensitivity data. Similarly, it has been shown that one can predict PTMs from the underlying DNA primary sequence. Results In this study, we introduce a deep learning architecture called DeepPTM for predicting histone PTMs from transcription factor binding data and the primary DNA sequence. Extensive experimental results show that our deep learning model outperforms the prediction accuracy of the model proposed in Benveniste et al. (PNAS 2014) and DeepHistone (BMC Genomics 2019). The competitive advantage of our framework lies in the synergistic use of deep learning combined with an effective pre-processing step. Our classification framework has also enabled the discovery that the knowledge of a small subset of transcription factors (which are histone-PTM and cell-type-specific) can provide almost the same prediction accuracy that can be obtained using all the transcription factors data. Availabilityand implementation https://github.com/dDipankar/DeepPTM. Supplementary information Supplementary data are available at Bioinformatics online.

Baisya, Dipankar Ranjan (ORCID:0000000267847359)↗
DOE OSTI2021

Improving deep learning-based protein distance prediction in CASP14

Abstract Motivation Accurate prediction of residue–residue distances is important for protein structure prediction. We developed several protein distance predictors based on a deep learning distance prediction method and blindly tested them in the 14th Critical Assessment of Protein Structure Prediction (CASP14). The prediction method uses deep residual neural networks with the channel-wise attention mechanism to classify the distance between every two residues into multiple distance intervals. The input features for the deep learning method include co-evolutionary features as well as other sequence-based features derived from multiple sequence alignments (MSAs). Three alignment methods are used with multiple protein sequence/profile databases to generate MSAs for input feature generation. Based on different configurations and training strategies of the deep learning method, five MULTICOM distance predictors were created to participate in the CASP14 experiment. Results Benchmarked on 37 hard CASP14 domains, the best performing MULTICOM predictor is ranked 5th out of 30 automated CASP14 distance prediction servers in terms of precision of top L/5 long-range contact predictions [i.e. classifying distances between two residues into two categories: in contact (<8 Angstrom) and not in contact otherwise] and performs better than the best CASP13 distance prediction method. The best performing MULTICOM predictor is also ranked 6th among automated server predictors in classifying inter-residue distances into 10 distance intervals defined by CASP14 according to the precision of distance classification. The results show that the quality and depth of MSAs depend on alignment methods and sequence databases and have a significant impact on the accuracy of distance prediction. Using larger training datasets and multiple complementary features improves prediction accuracy. However, the number of effective sequences in MSAs is only a weak indicator of the quality of MSAs and the accuracy of predicted distance maps. In contrast, there is a strong correlation between the accuracy of contact/distance predictions and the average probability of the predicted contacts, which can therefore be more effectively used to estimate the confidence of distance predictions and select predicted distance maps. Availability and implementation The software package, source code and data of DeepDist2 are freely available at https://github.com/multicom-toolbox/deepdist and https://zenodo.org/record/4712084#.YIIM13VKhQM. Supplementary information Supplementary data are available at Bioinformatics online.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

A deep dilated convolutional residual network for predicting interchain contacts of protein homodimers

Abstract Motivation Deep learning has revolutionized protein tertiary structure prediction recently. The cutting-edge deep learning methods such as AlphaFold can predict high-accuracy tertiary structures for most individual protein chains. However, the accuracy of predicting quaternary structures of protein complexes consisting of multiple chains is still relatively low due to lack of advanced deep learning methods in the field. Because interchain residue–residue contacts can be used as distance restraints to guide quaternary structure modeling, here we develop a deep dilated convolutional residual network method (DRCon) to predict interchain residue–residue contacts in homodimers from residue–residue co-evolutionary signals derived from multiple sequence alignments of monomers, intrachain residue–residue contacts of monomers extracted from true/predicted tertiary structures or predicted by deep learning, and other sequence and structural features. Results Tested on three homodimer test datasets (Homo_std dataset, DeepHomo dataset and CASP-CAPRI dataset), the precision of DRCon for top L/5 interchain contact predictions (L: length of monomer in a homodimer) is 43.46%, 47.10% and 33.50% respectively at 6 Å contact threshold, which is substantially better than DeepHomo and DNCON2_inter and similar to Glinter. Moreover, our experiments demonstrate that using predicted tertiary structure or intrachain contacts of monomers in the unbound state as input, DRCon still performs well, even though its accuracy is lower than using true tertiary structures in the bound state are used as input. Finally, our case study shows that good interchain contact predictions can be used to build high-accuracy quaternary structure models of homodimers. Availability and implementation The source code of DRCon is available at https://github.com/jianlin-cheng/DRCon. The datasets are available at https://zenodo.org/record/5998532#.YgF70vXMKsB. Supplementary information Supplementary data are available at Bioinformatics online.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

Genomic prediction of hybrid performance for agronomic traits in sorghum

Hybrid breeding in sorghum [Sorghum bicolor (L.) Moench] utilizes the cytoplasmic-nuclear male sterility (CMS) system for seed production and subsequently harnesses heterosis. Since the cost of developing and evaluating inbred and hybrid lines in the CMS system is costly and time-consuming, genomic prediction of parental lines and hybrids is based on genetic data genotype. We generated 602 hybrids by crossing two female (A) lines with 301 diverse and elite male (R) lines from the sorghum association panel and collected phenotypic data for agronomic traits over two years. We genotyped the inbred parents using whole genome resequencing and used 2,687,342 high quality (minor allele frequency > 2%) single nucleotide polymorphisms for genomic prediction. For grain yield, the experimental hybrids exhibited an average mid-parent heterosis of 40%. Genomic best linear unbiased prediction (GBLUP) for hybrid performance yielded an average prediction accuracy of 0.76–0.93 under the prediction scenario where both parental lines in validation sets were included in the training sets (T2). However, when only female tester was shared between training and validation sets (T1F), prediction accuracies declined by 12–90%, with plant height showing the greatest decline. Mean accuracies for predicting the general combining ability of male parents ranged from 0.33 to 0.62 for all traits. Our results showed hybrid performance for agronomic traits can be predicted with high accuracy, and optimizing genomic relationship is essential for optimal training population design for genomic selection in sorghum breeding.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2022

Evaluating the location capabilities of a regional infrasonic network in Utah, US, using both ray tracing-derived and empirical-derived celerity-range and backazimuth models

SUMMARY More realistic models for infrasound signal propagation across a region can be used to improve the precision and accuracy of spatial and temporal source localization estimates. Motivated by incomplete infrasound event bulletins in the Western US, the location capabilities of a regional infrasonic network of stations located between 84–458 km from the Utah Test and Training Range, Utah, USA, is assessed using a series of near-surface explosive events with complementary ground truth (GT) information. Signal arrival times and backazimuth estimates are determined with an automatic F-statistic based signal detector and manually refined by an analyst. This study represents the first application of three distinct celerity-range and backazimuth models to an extensive suite of realistic signal detections for event location purposes. A singular celerity and backazimuth deviation model was previously constructed using ray tracing analysis based on an extensive archive of historical atmospheric specifications and is applied within this study to test location capabilities. Similarly, a set of multivariate, season and location specific models for celerity and backazimuth are compared to an empirical model that depends on the observations across the infrasound network and the GT events, which accounts for atmospheric propagation variations from source to receiver. Discrepancies between observed and predicted signal celerities result in locations with poor accuracy. Application of the empirical model improves both spatial localization precision and accuracy; all but one location estimates retain the true GT location within the 90 per cent confidence bounds. Average mislocation of the events is 15.49 km and average 90 per cent error ellipse areas are 4141 km2. The empirical model additionally reduces origin time residuals; origin time residuals from the other location models are in excess of 160 s while residuals produced with the empirical model are within 30 s of the true origin time. We demonstrate that event location accuracy is driven by a combination of signal propagation model and the azimuthal gap of detecting stations. A direct relationship between mislocation, error ellipse area and increased station azimuthal gaps indicate that for sparse networks, detection backazimuths may drive location biases over traveltime estimates.

58 GEOSCIENCES↗
DOE OSTI2021

DeepMerge – II. Building robust deep learning algorithms for merging galaxy identification across domains

In astronomy, neural networks are often trained on simulation data with the prospect of being used on telescope observations. Unfortunately, training a model on simulation data and then applying it to instrument data leads to a substantial and potentially even detrimental decrease in model accuracy on the new target dataset. Simulated and instrument data represent different data domains, and for an algorithm to work in both, domain-invariant learning is necessary. Here we employ domain adaptation techniques— Maximum Mean Discrepancy (MMD) as an additional transfer loss and Domain Adversarial Neural Networks (DANNs)— and demonstrate their viability to extract domain-invariant features within the astronomical context of classifying merging and non-merging galaxies. Additionally, we explore the use of Fisher loss and entropy minimization to enforce better in-domain class discriminability. We show that the addition of each domain adaptation technique improves the performance of a classifier when compared to conventional deep learning algorithms. We demonstrate this on two examples: between two Illustris-1 simulated datasets of distant merging galaxies, and between Illustris-1 simulated data of nearby merging galaxies and observed data from the Sloan Digital Sky Survey. The use of domain adaptation techniques in our experiments leads to an increase of target domain classification accuracy of up to ~20%. With further development, these techniques will allow astronomers to successfully implement neural network models trained on simulation data to efficiently detect and study astrophysical objects in current and future large-scale astronomical surveys.

galaxies: interactions↗
DOE OSTI2023

Star cluster classification using deep transfer learning with PHANGS- HST

Currently available star cluster catalogues from the Hubble Space Telescope (HST) imaging of nearby galaxies heavily rely on visual inspection and classification of candidate clusters. Here, the time-consuming nature of this process has limited the production of reliable catalogues and thus also post-observation analysis. To address this problem, deep transfer learning has recently been used to create neural network models that accurately classify star cluster morphologies at production scale for nearby spiral galaxies (D ≲ 20 Mpc). Here, we use HST ultraviolet (UV)–optical imaging of over 20 000 sources in 23 galaxies from the Physics at High Angular resolution in Nearby GalaxieS (PHANGS) survey to train and evaluate two new sets of models: (i) distance-dependent models, based on cluster candidates binned by galaxy distance (9–12, 14–18, and 18–24 Mpc), and (ii) distance-independent models, based on the combined sample of candidates from all galaxies. We find that the overall accuracy of both sets of models is comparable to previous automated star cluster classification studies (~60–80 per cent) and shows improvement by a factor of 2 in classifying asymmetric and multipeaked clusters from PHANGS-HST. Somewhat surprisingly, while we observe a weak negative correlation between model accuracy and galactic distance, we find that training separate models for the three distance bins does not significantly improve classification accuracy. We also evaluate model accuracy as a function of cluster properties such as brightness, colour, and spectral energy distribution (SED)-fit age. Based on the success of these experiments, our models will provide classifications for the full set of PHANGS-HST candidate clusters (N ~ 200 000) for public release.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2024

Towards modelling AR Sco: generalized particle dynamics and strong radiation-reaction regimes

Numerical simulations of relativistic plasmas have become more feasible, popular, and crucial for various astrophysical sources with the availability of computational resources. The necessity for high-accuracy particle dynamics is especially highlighted in pulsar modelling due to the extreme associated electromagnetic fields and particle Lorentz factors. Including the radiation-reaction force in the particle dynamics adds even more complexity to the problem, but is crucial for such extreme astrophysical sources. We have also realized the need for such modelling concerning magnetic mirroring and particle injection models proposed for AR Sco, the first white dwarf pulsar. This paper demonstrates the benefits of using higher-order explicit numerical integrators with adaptive time-step methods to solve the full particle dynamics with radiation-reaction forces included. We show that for standard test scenarios, namely various combinations of uniform E- and B-fields and a static dipole B-field, the schemes we use are equivalent to and in extreme field cases outperform standard symplectic integrators in accuracy. We show that the higher-order schemes have massive computational time improvements due to the adaptive time-steps we implement, especially in non-uniform field scenarios and included radiation reaction where the particle gyro-radius rapidly changes. When balancing accuracy and computational time, we identified the adaptive Dormand–Prince eighth-order scheme to be ideal for our use cases. The schemes we use maintain accuracy and stability in describing the particle dynamics and we indicate how a charged particle enters radiation-reaction equilibrium and conforms to the analytical Aristotelian Electrodynamics expectations.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2019

Measurements and calculations of 2 s – 2 p transitions in neonlike germanium: Achieving agreement at the 10 – 4 level

Calculations that achieve spectroscopic accuracy, i.e., at the level of 10 –4 or better, are rare and typically restricted to few-electron systems. Here we present high-resolution grating spectrometer measurements of 11 2s 1/2 – 2p 3/2 transitions in neonlike Ge 22+ that distinguish between two different implementations of the many-body perturbation theory (MBPT) method, including our own. The measurements show that our implementation of the MBPT method consistently achieves spectroscopic accuracy at the 10 –4 level for all of the measured lines. Moreover, our calculations improve the 10 –3 accuracy of prior MBPT calculations by an order of magnitude and allow us to make greatly improved predictions of the location of potential photopumped lines in Ge 22+ . Here, the measurements also improve on previous measurements of 2s 1/2 – 2p 3/2 transitions in fluorinelike Ge 23+ and oxygenlike Ge 24+ . Measurements of n = 2 to n = 3 transitions in Al 7+ , Al 8+ , and Al 9+ are used to support our measurement accuracy at the 10 –5 level.

74 ATOMIC AND MOLECULAR PHYSICS↗
DOE OSTI2021

Lattice Boltzmann model for conjugate heat transfer across thin walls

A lattice Boltzmann (LB) model with an efficient and accurate interface treatment for conjugate heat transfer across a thin wall between two different media is developed. The proposed interface treatment avoids fine meshing and computation within the thin layer; instead, the energy balance within the thin layer and the conjugate conditions on each interface are utilized to construct explicit updating schemes for the microscopic distribution functions of the LB model at the interior lattice nodes of the two media next to the thin layer. The proposed interface schemes reduce to the standard interface scheme for conjugate conditions in the literature in the limit of zero thickness of the thin layer, and thus it can be considered a more general interface treatment. A simplified version of the interface treatment is also proposed when the heat flux variation along the tangential direction of the thin layer is negligible. Three representative numerical tests are conducted to verify the applicability and accuracy of the proposed interface schemes. The results demonstrate that the intrinsic second-order accuracy of the LB model is preserved with the proposed interface schemes for thin layers with constant tangential fluxes, while for general situations with varying tangential fluxes, first-order accuracy is obtained. Finally, this interface treatment within the LB framework is attractive in conjugate heat transfer modeling involving thin layers for its simplicity, accuracy, and significant reduction in computational resources.

42 ENGINEERING↗
DOE OSTI2025

A Near-Real-Time Model for Predicting Electricity Disruptions in Texas During Winter Storms

There has been an increase in extreme weather events, posing a threat to power grid systems, potentially influenced by factors such as population growth, changes in ecosystems, land cover, and land use in the service area, as well as the growth of certain vegetation types. This research seeks to develop a predictive model to mitigate potential damages caused by future winter storms. This research utilizes the Light Gradient Boosting Machine (LightGBM), incorporating the number of power outages experienced at the county level, geographic details, weather information, and lagged outage and lagged weather data. The developed models were broadly divided into two groups, with six models in each group - one group without optimization and another with optimization, totaling 12 trained models. For model optimization, Bayesian optimization was employed using Root Mean Squared Error (RMSE) as the objective function. In results, when comparing Group 2 (the optimized group) with Group 1 (the non-optimized group), it was found that optimization did not always lead to a reduction in RMSE and Mean Absolute Error (MAE). However, in terms of Mean Directional Accuracy (MDA), while all results in Group 1 were below the baseline accuracy of 0.33, all results in Group 2 exceeded 0.33, with some cases showing an increase of more than three times the baseline. The results indicated that, in the optimized model group, Population and Pressure were the most influential factors when using current weather data and geographical information. When using lagged data, lagged recorded outages and lagged Pressure emerged as the most significant factors. Among the 12 developed models, the L-1-2-O model showed the lowest RMSE and MAE, as well as the highest accuracy, with values of 390.62 households and 168.13 households, respectively. To normalize the RMSE and MAE values, each metric was divided by the average number of households among the counties in Texas. For the L-1-2-O model, the scaled RMSE was 0.88% and the scaled MAE was 0.38%. In terms of MDA, which indicates the accuracy of the prediction direction, the L-1-O model achieved the highest score of 0.41. Although this study focused on Texas, which suffered the greatest impact from the winter storms in 2021, with additional validation, the methodology used in this research could be applied to other regions.

Lee, Jangjae [Texas A & M Univ., College Station, ↗
DOE OSTI2026

Real-Time GPU-Accelerated OFDR With an Integrated Auxiliary Interferometer

A GPU-accelerated optical frequency domain reflectometry (OFDR) system with an improved integrated auxiliary interferometer is proposed. Unlike conventional approaches that require separate auxiliary interferometers and multiple detection channels, the proposed OFDR system embeds this functionality directly into the signal via an intentional beat component. This enables self-calibration of laser nonlinearity while maintaining a cost-effective hardware configuration. Building on this simplified configuration, the system leverages GPU acceleration with an NVIDIA RTX 4070 Ti to achieve real-time performance, delivering high-throughput signal processing for continuous OFDR interrogation. The signal processing pipeline comprises signal capture, resampling for nonlinearity compensation, and frequency shift computation, all optimized for parallel execution. Hardware benchmarking demonstrates substantial acceleration over CPU implementations, achieving up to a 45× speedup for resampling and frequency shift computations and enabling processing latencies below 30 ms. Thermal response validation is conducted under two complementary scenarios: localized heating using a water bath and cryogenic-temperature conditions using liquid nitrogen. Under localized heating, the system achieves an accuracy of 0.249 °C with a thermal sensitivity of 5.971 GHz/°C, while cryogenic-temperature validation demonstrates a frequency shift response with a sensitivity of 2.383 GHz/°C and an accuracy of 2.04 °C. The high acceleration of the proposed GPU-accelerated OFDR system and its accuracy are achieved by exploiting CUDA-based stride indexing, enabling efficient parallel segmentation and processing of large datasets without additional memory copies. The benchmarking results confirm the robustness, accuracy, and deployability of the proposed OFDR system across a wide temperature range, establishing it as a practical platform for real-time distributed fiber sensing in structurally dynamic environments.

Harb, Salah [Lawrence Berkeley National Laboratory↗
DOE OSTI2022

VWC-BERT: Scaling Vulnerability–Weakness–Exploit Mapping on Modern AI Accelerators

Defending cybersystems needs accurate mapping of software and hardware vulnerabilities to generalized descriptions of weaknesses, and weaknesses to exploits. These mappings enable cyber defenders to build plans for effective defense and assessment of potential risks to a cybersystem. With close to 170k vulnerabilities, manual mapping is not a feasible option. However, automated mapping is challenging due to limited training data, computational intractability, and limitations in computational natural language processing. Tools based on breakthroughs in Transformer-based language models have been demonstrated to classify vulnerabilities with high accuracy. We make three key contributions in this paper: (1) We present a new framework, \VWCBERT, that augments the Transformer-based hierarchical multi-class classification framework of Das et al. (\textsc{V2W-BERT}) with the ability to map weaknesses to exploits. (2) We implement \VWCBERT~ on modern AI accelerator platforms using two data parallel techniques for the pre-training phase and demonstrate nearly linear speedups across NVIDIA and Graphcore accelerator platforms. We observe nearly linear speedups for up to 16 V100 and 8 A100 GPUs, and about 3.4$\times$ speedup for A100 relative to V100 GPUs. We also observe excellent speedups on Graphcore, with $5.7\times$ speedup on 128 IPUs relative to 16 IPUs. Enabled by scaling, we also demonstrate higher accuracy using a larger language model, RoBERTa-Large. We show up to 87\% accuracy for strict and up to 98\% accuracy for relaxed classification. (3) We develop a novel parallel link manager for the link prediction phase and demonstrate up to 21$\times$ speedup with 16 V100 GPUs relative to one V100 GPU, and thus reducing the runtime from 2.5 hours to 10 minutes. We believe that generalizability and scalability of \VWCBERT~ will benefit both the theoretical development and practical deployment of novel cyberdefense solutions and vulnerability classification.

Das, Siddhartha Shankar↗
DOE OSTI2022

New Hybrid Model for Evaluating the Frequency-Dependent Leakage Inductance of a Variable Inductance Transformer (VIT)

Skin and proximity effects can cause a significant drop in the effective leakage inductance of a transformer when the operating frequency is increased. Although the magnetic image method-based double-2-D model can calculate the lowfrequency leakage inductance with sufficient accuracy, it is inherently a frequency-independent model. While Dowell’s 1-D model uses frequency-dependent relations to account for both skin and proximity effects, its accuracy is severely affected by the assumed winding geometry. In this paper, a hybrid model is proposed that uses superposition to combine a modified Dowell’s model with the double-2-D model. The proposed model is investigated on a variable inductance transformer (VIT)—a partially-filled transformer whose leakage inductance can be varied by moving one of the windings mechanically. The frequency-dependent leakage inductances of the VIT evaluated using the hybrid model are in excellent agreement with the corresponding finite element method (FEM) simulated and experimentally measured values, thereby validating the proposed hybrid model.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2023

Accelerating Scientific Simulations with Bi-Fidelity Weighted Transfer Learning

High-fidelity modeling is an essential design tool for many engineering applications. However, for complex systems, computational cost can be a limiting factor. Analyzing parameter sensitivity, uncertainty quantification, and design optimization require many model evaluations. Surrogate models are often used to develop the relationship between model parameters and quantities of interest. However, in the case of complex systems, surrogate models require several degrees of freedom and, thus, a large number of data points to determine the correct dependencies. For many applications, this may be prohibitively expensive. The reduction of computational requirements can be achieved by leveraging low-fidelity models. Low-fidelity models represent the system at a coarser resolution with the advantage of computational efficiency. Therefore, a bi-fidelity modeling paradigm, which augments the accuracy of a low-fidelity model in a computationally efficient manner by invoking limited runs of a high-fidelity model, can be leveraged to sufficiently balance the accuracy and computational requirements. In this work, a bi-fidelity weighted transfer learning method using neural networks was applied to a computational fluid dynamics heat transfer modeling problem. The transfer learning advantage was investigated as a function of hyperparameters. Our main finding is that the use of a bi-fidelity modeling paradigm achieves accuracy close to that of a high-fidelity Gaussian process model while significantly reducing computational cost. The bi-fidelity model achieves comparable performance with 90 high-fidelity samples-that is, 60% less than the samples needed to achieve similar accuracy without the use of bi-fidelity modeling,

Borowiec, Katarzyna↗
DOE OSTI2026

Interpretable Models for Workflow Differentiation in High-Performance Scientific Networks

Scientific workflows in high-performance networks spawn hundreds of interdependent flows that must be managed collectively—yet existing network classifiers treat each flow in isolation, leading to fragmented QoS decisions and missed interflow patterns. We present a novel traffic classification solution that operates at the workflow level, distinguishing entire filetransfer operations from streaming analytics by capturing how concurrent flows interact and burst together. We introduce a workflow identification window (WIW) that ingests raw packet headers from parallel flows into unified tensors, preserving the spatial-temporal patterns that differentiate scientific workflows. This approach achieves 98.7% accuracy using CNN, LSTM, and hybrid architectures, while maintaining 84% accuracy on production traffic collected a week later—demonstrating robustness to temporal drift. By integrating SHAP and GradCAM explainability, we reveal that early-packet timing patterns and cross-flow correlations drive classification decisions, providing operators with interpretable insights. Our system enables coherent workflow-level QoS enforcement and dynamic bandwidth allocation in scientific networks, eliminating manual per-flow configuration while maintaining classification latency at millisecond level.

Giannakou, Anna [LBL, Berkeley]↗