Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “model weights”

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 91 records · Page 5
All sourcesNASA NTRSDOE OSTI
NASA NTRS1992

Mechanical design of a rotary balance system for NASA. Langley Research Center's vertical spin tunnel

A new, lightweight Rotary Balance System is presently being fabricated and installed as part of a major upgrade to the existing 20-Foot Vertical Spin Tunnel (Figure 1) located at NASA's Langley Research Center. This upgrade, to improve model testing productivity of the only free-spinning vertical wind tunnel in the Western Hemisphere, includes a modern fan/drive and tunnel control system, an updated video recording system, and the new rotary balance system. The rotary balance is a mechanical apparatus which enables the measurement of aerodynamic force and moment data under spinning conditions (100 rpm). This data is used in spin analysis and is vital to the implementation of large amplitude maneuvering simulations required for all new high performance aircraft. The present system is over 40 years old and is worn beyond acceptable tolerances, as well as being susceptible to frequent breakdowns. It has serious limitations in model weight, position adjustment and spin-rate capabilities. The new rotary balance system described in this report will permit greater test efficiency and improved data accuracy. Rotary Balance testing with the model enclosed in a tare bag can also be performed to obtain resulting model forces from the spinning operation. The rotary balance system will be stored against the tunnel sidewall during free-flight model testing.

J. W. Allred↗
NASA NTRS1994

Radiative transfer in shrub savanna sites in Niger: Preliminary results from HAPEX-Sahel. Part 1: Modelling surface reflectance using a geometric-optical approach

To use optical remote sensing to monitor land surface-climate interactions over large areas, algorithms must be developed to relate multispectral measurements to key variables controlling the exchange of matter (water, carbon dioxide) and energy between the land surface and the atmosphere. The proportion of the ground covered by vegetation and the interception of photosynthetically active radiation (PAR) by vegetation are examples of two variables related to evapotranspiration and primary production, respectively. An areal-proportion model of the multispectral reflectance of shrub savanna, composed of scattered shrubs with a grass, forb or soil understory, predicted the reflectance of two 0.5 km(exp 2) sites as the area-weighted average of the shrub and understory or 'background' reflectances. Although the shaded crown and shaded background have darker reflectances, ignoring them in the area-weighted model is not serious when shrub cover is low and solar zenith angle is small. A submodel predicted the reflectance of the shrub crown as a function of the foliage reflectance and amount of plant material within the crown, and the background reflectance scattered or transmitted through canopy gaps (referred to as a soil-plant 'spectral interaction' term). One may be able to combine these two models to estimate both the fraction of vegetation cover and interception of PAR by green vegetation in a shrub savanna.

Franklin, Janet↗
DOE OSTI2023

Directionally supervised cellular automaton for the initial peopling of Sahul

Reconstructing the patterns of Homo sapiens expansion out of Africa and across the globe has been advanced using demographic and travel-cost models. However, modelled routes are ipso facto influenced by migration rates, and vice versa. We combined movement ‘superhighways’ with a demographic cellular automaton to predict one of the world's earliest peopling events — Sahul between 75000 and 50000 years ago. Novel outcomes from the superhighways-weighted model include (i) an approximate doubling of the predicted time to continental saturation (~10,000 years) compared to that based on the directionally unsupervised model (~5000 years), suggesting that rates of migration need to account for topographical constraints in addition to rate of saturation; (ii) a previously undetected movement corridor south through the centre of Sahul early in the expansion wave based on the scenarios assuming two dominant entry points into Sahul; and (iii) a better fit to the spatially de-biased, Signor-Lipps-corrected layer of initial arrival inferred from dated archaeological material. Our combined model infrastructure provides a data-driven means to examine how people initially moved through, settled, and abandoned different regions of the globe.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗
NASA NTRS1998

A Conceptual Approach to Assimilating Remote Sensing Data to Improve Soil Moisture Profile Estimates in a Surface Flux/Hydrology Model: Disaggregation - 3

This is a followup on the preceding presentation by Crosson and Schamschula. The grid size for remote microwave measurements is much coarser than the hydrological model computational grids. To validate the hydrological models with measurements we propose mechanisms to disaggregate the microwave measurements to allow comparison with outputs from the hydrological models. Weighted interpolation and Bayesian methods are proposed to facilitate the comparison. While remote measurements occur at a large scale, they reflect underlying small-scale features. We can give continuing estimates of the small scale features by correcting the simple 0th-order, starting with each small-scale model with each large-scale measurement using a straightforward method based on Kalman filtering.

Caulfield, John↗
DOE OSTI2024

Dataset for "Evaluating Deep Learning Approaches for Predictions in Unmonitored Basins with Continental-scale Stream Temperature Models" Willard et al. (2024)

This data release provides all data and code used in the paper " "Evaluating Deep Learning Approaches for Predictions in Unmonitored Basins with Continental-scale Stream Temperature Models" Willard et al. (2024)" to model stream temperature, evaluate, and assess results. The associated manuscript explores current open questions in prediction in ungauged and unmonitored basins concerning top-down versus bottom-up approaches, tradeoffs between data available and input requirements, and the appropriate representation of catchment attributes as inputs to deep learning models. Modeling was done primarily with long short-term memory (LSTM) models, and stream site coverage spans 1362 locations across the conterminous United States. The data is organized into these items items:Code repository and data for the paper " "Evaluating Deep Learning Approaches for Predictions in Unmonitored Basins with Continental-scale Stream Temperature Models" Willard et al. (2024)".Code: stream_temp_ml_regionalization.zip contains the code repositoryData to run the code: - data_dir.zip -- contains all files that should be moved to the "DATA_DIR" variable defined in the "set_env_vars.sh" script in the code repository- metadata_dir.zip -- contains all files that should be moved to the "METADATA_DIR" variable defined in the "set_env_vars.sh" script in the code repository- error_analysis_attribute_and_groundwater_dir.zip - workflows for the extended error analysis by stream attribute and groundwater influenceData produced by the code and used in the paper:- outputs_dir.zip - contains model output and results (outputs_dir/results), model weights (outputs_dir/models), and all other outputs used for the paper including feature importances.To cite this code, please use the following BibTeX or MLA entries:bibtex:@misc{willard2024streamdata, author = {Jared Willard and Fabio Ciulla and Helen Weierbach and Vipin Kumar and Charuleka Varadharajan}, title = {Dataset for "Evaluating Deep Learning Approaches for Predictions in Unmonitored Basins with Continental-scale Stream Temperature Models"}, year = {2024}, doi = {10.15485/2448016}, publisher = {ESS-DIVE Repository}, url = {https://doi.org/10.15485/2448016}}MLA: Willard, Jared, et al. Dataset for "Evaluating Deep Learning Approaches for Predictions in Unmonitored Basins with Continental-scale Stream Temperature Models". 2024. ESS-DIVE Repository, doi:10.15485/2448016.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2014

Effect of cosolvent on protein stability: A theoretical investigation

We developed a statistical thermodynamic algorithm for analyzing solvent-induced folding/unfolding transitions of proteins. The energetics of protein transitions is governed by the interplay between the cavity formation contribution and the term reflecting direct solute-cosolvent interactions. The latter is viewed as an exchange reaction in which the binding of a cosolvent to a solute is accompanied by release of waters of hydration to the bulk. Our model clearly differentiates between the stoichiometric and non-stoichiometric interactions of solvent or co-solvent molecules with a solute. We analyzed the urea- and glycine betaine (GB)-induced conformational transitions of model proteins of varying size which are geometrically approximated by a sphere in their native state and a spherocylinder in their unfolded state. The free energy of cavity formation and its changes accompanying protein transitions were computed based on the concepts of scaled particle theory. The free energy of direct solute-cosolvent interactions were analyzed using empirical parameters previously determined for urea and GB interactions with low molecular weight model compounds. Our computations correctly capture the mode of action of urea and GB and yield realistic numbers for (∂ΔG°/∂a{sub 3}){sub T,P} which are related to the m-values of protein denaturation. Urea is characterized by negative values of (∂ΔG°/∂a{sub 3}){sub T,P} within the entire range of urea concentrations analyzed. At concentrations below ∼1 M, GB exhibits positive values of (∂ΔG°/∂a{sub 3}){sub T,P} which turn positive at higher GB concentrations. The balance between the thermodynamic contributions of cavity formation and direct solute-cosolvent interactions that, ultimately, defines the mode of cosolvent action is extremely subtle. A 20% increase or decrease in the equilibrium constant for solute-cosolvent binding may change the sign of (∂ΔG°/∂a{sub 3}){sub T,P} thereby altering the mode of cosolvent action (stabilizing to destabilizing or vice versa)

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

SymbolNet: neural symbolic regression with adaptive dynamic pruning for compression

Abstract Compact symbolic expressions have been shown to be more efficient than neural network (NN) models in terms of resource consumption and inference speed when implemented on custom hardware such as field-programmable gate arrays (FPGAs), while maintaining comparable accuracy (Tsoi et al 2024 EPJ Web Conf. 295 09036). These capabilities are highly valuable in environments with stringent computational resource constraints, such as high-energy physics experiments at the CERN Large Hadron Collider. However, finding compact expressions for high-dimensional datasets remains challenging due to the inherent limitations of genetic programming (GP), the search algorithm of most symbolic regression (SR) methods. Contrary to GP, the NN approach to SR offers scalability to high-dimensional inputs and leverages gradient methods for faster equation searching. Common ways of constraining expression complexity often involve multistage pruning with fine-tuning, which can result in significant performance loss. In this work, we propose S y m b o l N e t , a NN approach to SR specifically designed as a model compression technique, aimed at enabling low-latency inference for high-dimensional inputs on custom hardware such as FPGAs. This framework allows dynamic pruning of model weights, input features, and mathematical operators in a single training process, where both training loss and expression complexity are optimized simultaneously. We introduce a sparsity regularization term for each pruning type, which can adaptively adjust its strength, leading to convergence at a target sparsity ratio. Unlike most existing SR methods that struggle with datasets containing more than O ( 10 ) inputs, we demonstrate the effectiveness of our model on the LHC jet tagging task (16 inputs), MNIST (784 inputs), and SVHN (3072 inputs).

Tsoi, Ho Fung (ORCID:0000000225502184)↗
DOE OSTI2019

Quantizing a solitonic string

Quite often the zero mode dynamics on solitonic vortices are described by a non-conformal effective world-sheet sigma model (WSSM). We address the problem of solitonic string quantization in this case. As well-known, only critical strings with conformal WSSMs are self-consistent in ultra-violet (UV) domain. Thus, we look for the appropriate UV completion of the low-energy non-conformal WSSM. We argue that for the solitonic strings supported in well-defined bulk theories the UV complete WSSM has a UV fixed point which can be used for string quantization. As an example, we consider BPS non- Abelian vortices supported by four-dimensional (4D) N = 2 SQCD with the gauge group U(N) and Nf quark multiplets where N f ≥ N. In addition to translational moduli the non-Abelian vortex under consideration carries orientational and size moduli. Their low- energy dynamics are described by a two-dimensional N = (2, 2) supersymmetric weighted model, namely, WCP (N, N f – N ). Given our UV completion of this WSSM we find its UV fixed point. The latter defines a superconformal WSSM. We observe two cases in which this conformal WSSM, combined with the free theory for four translational moduli, has ten-dimensional target space required for superstrings to be critical.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
NASA NTRS2012

Spinoff 2011

Topics include: Bioreactors Drive Advances in Tissue Engineering; Tooling Techniques Enhance Medical Imaging; Ventilator Technologies Sustain Critically Injured Patients; Protein Innovations Advance Drug Treatments, Skin Care; Mass Analyzers Facilitate Research on Addiction; Frameworks Coordinate Scientific Data Management; Cameras Improve Navigation for Pilots, Drivers; Integrated Design Tools Reduce Risk, Cost; Advisory Systems Save Time, Fuel for Airlines; Modeling Programs Increase Aircraft Design Safety; Fly-by-Wire Systems Enable Safer, More Efficient Flight; Modified Fittings Enhance Industrial Safety; Simulation Tools Model Icing for Aircraft Design; Information Systems Coordinate Emergency Management; Imaging Systems Provide Maps for U.S. Soldiers; High-Pressure Systems Suppress Fires in Seconds; Alloy-Enhanced Fans Maintain Fresh Air in Tunnels; Control Algorithms Charge Batteries Faster; Software Programs Derive Measurements from Photographs; Retrofits Convert Gas Vehicles into Hybrids; NASA Missions Inspire Online Video Games; Monitors Track Vital Signs for Fitness and Safety; Thermal Components Boost Performance of HVAC Systems; World Wind Tools Reveal Environmental Change; Analyzers Measure Greenhouse Gasses, Airborne Pollutants; Remediation Technologies Eliminate Contaminants; Receivers Gather Data for Climate, Weather Prediction; Coating Processes Boost Performance of Solar Cells; Analyzers Provide Water Security in Space and on Earth; Catalyst Substrates Remove Contaminants, Produce Fuel; Rocket Engine Innovations Advance Clean Energy; Technologies Render Views of Earth for Virtual Navigation; Content Platforms Meet Data Storage, Retrieval Needs; Tools Ensure Reliability of Critical Software; Electronic Handbooks Simplify Process Management; Software Innovations Speed Scientific Computing; Controller Chips Preserve Microprocessor Function; Nanotube Production Devices Expand Research Capabilities; Custom Machines Advance Composite Manufacturing; Polyimide Foams Offer Superior Insulation; Beam Steering Devices Reduce Payload Weight; Models Support Energy-Saving Microwave Technologies; Materials Advance Chemical Propulsion Technology; and High-Temperature Coatings Offer Energy Savings.

Source record↗
DOE OSTI2026

A comparative study of multimodal data fusion strategies for planetary spectroscopy

Integrating heterogeneous data sources can improve scientific inference when different modalities capture complementary information, but doing so is challenging in high-dimensional, small-sample settings. In spectroscopy for planetary exploration, Laser-Induced Breakdown Spectroscopy (LIBS), Raman Spectroscopy (Raman), Visible Infrared Spectroscopy (VISIR), and Mid-Infrared Spectroscopy (MIR) each examine different aspects of composition and mineralogy, raising fundamental questions about when and how data fusion improves predictive performance. Using a Mars-relevant set of geologic standards with measurements from all four modalities, we present a rigorous systematic evaluation of four data fusion strategies: low-level (data) fusion, mid-level (feature) fusion, high-level (decision) fusion, and residual-boosting (sequential) fusion. We assess performance in predicting oxide composition via nested cross-validation and corrected significance testing to evaluate whether data fusion improves upon single-modality baselines. We show that data fusion does not uniformly improve accuracy, and that observed gains are modest, oxide-dependent, and sensitive to modality and model structure. To move beyond aggregate accuracy metrics, we use model coefficients, permutation importance, and residual gain analysis to examine how the fusion models weight individual modalities and to identify patterns of apparent complementarity or redundancy. Though focused on spectroscopy for planetary exploration, our framework for data fusion evaluation and interpretation extends to other scientific domains with heterogeneous and scarce data and provides a principled approach evaluating data fusion strategies, interpreting modality contributions, and understanding tradeoffs among data fusion strategies.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2015

Total Laryngectomy Versus Larynx Preservation for T4a Larynx Cancer: Patterns of Care and Survival Outcomes

Purpose: To examine practice patterns and compare survival outcomes between total laryngectomy (TL) and larynx preservation chemoradiation (LP-CRT) in the setting of T4a larynx cancer, using a large national cancer registry. Methods and Materials: Using the National Cancer Database, we identified 969 patients from 2003 to 2006 with T4a squamous cell larynx cancer receiving definitive treatment with either initial TL plus adjuvant therapy or LP-CRT. Univariate and multivariable logistic regression were used to assess predictors of undergoing surgery. Survival outcomes were compared using Kaplan-Meier and propensity score–adjusted and inverse probability of treatment–weighted Cox proportional hazards methods. Sensitivity analyses were performed to account for unmeasured confounders. Results: A total of 616 patients (64%) received LP-CRT, and 353 (36%) received TL. On multivariable logistic regression, patients with advanced nodal disease were less likely to receive TL (N2 vs N0, 26.6% vs 43.4%, odds ratio [OR] 0.52, 95% confidence interval [CI] 0.37-0.73; N3 vs N0, 19.1% vs 43.4%, OR 0.23, 95% CI 0.07-0.77), whereas patients treated in high case-volume facilities were more likely to receive TL (46.1% vs 31.5%, OR 1.78, 95% CI 1.27-2.48). Median survival for TL versus LP was 61 versus 39 months (P<.001). After controlling for potential confounders, LP-CRT had inferior overall survival compared with TL (hazard ratio 1.31, 95% CI 1.10-1.57), and with the inverse probability of treatment–weighted model (hazard ratio 1.25, 95% CI 1.05-1.49). This survival difference was shown to be robust on additional sensitivity analyses. Conclusions: Most patients with T4a larynx cancer receive LP-CRT, despite guidelines suggesting TL as the preferred initial approach. Patients receiving LP-CRT had more advanced nodal disease and worse overall survival. Previous studies of (non-T4a) locally advanced larynx cancer showing no difference in survival between LP-CRT and TL may not apply to T4a disease, and patients should be counseled accordingly.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2016

X-ray and neutron total scattering analysis of Hy·(Bi0.2Ca0.55Sr0.25)(Ag0.25Na0.75)Nb3O10·xH2O perovskite nanosheet booklets with stacking disorder

Ion-exchanged Aurivillius materials form perovskite nanosheet booklets wherein well-defined bi-periodic sheets, with ~11.5 Å thickness, exhibit extensive stacking disorder. The perovskite layer contents were defined initially using combined synchrotron X-ray and neutron Rietveld refinement of the parent Aurivillius structure. The structure of the subsequently ion-exchanged material, which is disordered in its stacking sequence, is analyzed using both pair distribution function (PDF) analysis and recursive method simulations of the scattered intensity. Combined X-ray and neutron PDF refinement of supercell stacking models demonstrates sensitivity of the PDF to both perpendicular and transverse stacking vector components. Further, hierarchical ensembles of stacking models weighted by a standard normal distribution are demonstrated to improve PDF fit over 1–25 Å. Recursive method simulations of the X-ray scattering profile demonstrate agreement between the real space stacking analysis and more conventional reciprocal space methods. The local structure of the perovskite sheet is demonstrated to relax only slightly from the Aurivillius structure after ion exchange.

Metz, Peter↗
DOE OSTI2021

Energy consumption and charging load profiles from long-haul truck electrification in the United States

Abstract The urgent need to decarbonize the transportation sector combined with falling battery prices has spurred industry and policy interest in long-haul truck electrification. The charging behavior and resulting loads from electrified long-haul freight trucks are crucial for the smooth operation of the electric grid and have far-reaching environmental impacts (e.g., greenhouse gas and other air pollutant emissions). However, the aggregate energy impact of a fleetwide shift to electrified long-haul freight trucking has not been explored. This study combines electric truck design scenarios, bottom-up truck weight modeling, vehicle energy modeling, large-scale truck traffic data, and simulation of likely operation and charging behaviors to estimate end-use energy consumption and location-specific hourly charging loads for a national fleet of long-haul electric trucks. Relative to a fleet of future diesel trucks, electrification would reduce direct end-use energy consumption by 0.9 × 10 18 J (0.9 quadrillion BTU), but electrification might increase life cycle energy consumption depending on the electricity source. The electricity required to charge long-haul electric trucks is equivalent to five percent of annual electricity consumption in the United States (US). The simulated truck charging loads peak during the day across the US grid regions, but the charging peaks’ exact timing is sensitive to when trucks are dispatched for operation. The load shapes suggest that electric trucks’ charging loads can coincide with peaks in solar power generation, and planning could enable on- or off-site integration between truck charging stations and renewable electricity generation.

Tong, Fan (ORCID:0000000346613956)↗
DOE OSTI2026

FLYING SERVING: On-the-Fly Parallelism Switching for Large Language Model Serving

Production LLM serving must simultaneously deliver high throughput, low latency, and sufficient context capacity under non-stationary traffic and mixed request requirements. Data parallelism (DP) maximizes throughput by running independent replicas, while tensor parallelism (TP) reduces per-request latency and pools memory for long-context inference. However, existing serving stacks typically commit to a static parallelism configuration at deployment; adapting to bursts, priorities, or long-context requests is often disruptive and slow. We present Flying Serving, a vLLM-based system that enables online DP-TP switching without restarting engine workers. Flying Serving makes reconfiguration practical by virtualizing the state that would otherwise force data movement: (i) a zero-copy Model Weights Manager that exposes TP shard views on demand, (ii) a KV Cache Adaptor that preserves request KV state across DP/TP layouts, (iii) an eagerly initialized Communicator Pool to amortize collective setup, and (iv) a deadlock-free scheduler that coordinates safe transitions under execution skew. Across three popular LLMs and realistic serving scenarios, Flying Serving improves performance by up to 4.79 × under high load and 3.47 × under low load while supporting latency- and memory-driven requests.

Gao, Shouwei [ORNL]↗
DOE OSTI2019

Diet of Nestling Red-Headed Woodpeckers in South Carolina

Melanerpes erythrocephalus (Red-headed Woodpecker) has experienced sharp declines in portions of its range. Knowledge of how birds use their nesting habitats, particularly what foods they exploit, may be important in determining causes of population declines, but no modern quantitative study exists on diets of nestling Red-headed Woodpeckers. Our objectives were to identify diets of nestling Red-headed Woodpeckers and quantify variability in food types over time and between roles of males and females in provisioning their young. We conducted observations of nests on the Savannah River Site, SC, from June to September, 2006–2007. We recorded 791 food items fed to nestlings, representing 7 taxa of plants and 18 taxa of animals (16 invertebrate, 2 vertebrate). We assigned food items as either animal matter or soft mast and compared proportions using a binomial mixed model approach. Of the 12 models we tested, 3 received 67% of the cumulative AIC model weight and all included either year or month, indicating annual and monthly variation in foods fed to nestlings. Animal matter composed the majority of Red-headed Woodpecker nestling foods (71.5%), but notably, soft mast was an important component (28.5%). As a result, we suggest that future research on Red-headed Woodpeckers consider how the availability of soft mast may or may not limit productivity of this species.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2021

A Bayesian Neural Network Ensemble Approach for Improving Large-Scale Streamflow Predictability

Predictive modeling through the use of AI techniques and AI-derived model components; the use of AI and other tools to design a prediction system comprising of a hierarchy of models (e.g.,AI driven model/component/parameterization selection). We propose a Bayesian neural network ensemble approach for improving large-scale streamflow predictability and understanding in a changing environment that combines multiple Earth system land model predictions by calculating spatiotemporally varying model weights and biases while accounting for various types of observations at multiple scales with uncertainty.

58 GEOSCIENCES↗
DOE OSTI2025

Practical Probabilistic Programming

Recent advances in probabilistic programming languages (PPLs) have provided the capability for exact inference: computing a closed-form probability distribution for a given probabilistic program. In particular, the new language Roulette uses a language oriented programming (LOP) approach, wherein analysts build new programming languages on top of a set of primitives provided by Roulette, which then translates these structures into a weighted model counting problem which can be solved by automated reasoning tools. However, because Roulette provides few convenience features, developing these new languages is challenging even for expert users. We developed a standard library of common probability functions for Roulette with the goal of improved usability. This included approximation of continuous probability density functions using discrete probability mass functions. We demonstrated this approach by modeling a cosmic ray striking a RAM controller. We found that Roulette provides a powerful interface for highly expressive probabilistic programs to be generated. In collaboration with the NNSA Advanced Simulation and Computing program, which resulted in development of a tool called Circulette, we were able to model complex circuits expressed in Verilog using probabilistic programs with an expressivity not previously possible. Our research question that motivated the development of a Roulette standard library was to determine whether non-experts could use a PPL to model relevant problems regarding radiation effects on microelectronics. This standard library improved the expressivity of Roulette by implementing common probability density functions, mathematical operators on distributions, and support for empirical distributions. While Roulette is a powerful modeling language, the untyped, LOP approach makes error messages difficult to understand and requires expert aid. We recommend further research on Roulette, especially with its error messages, to enable improved usability. At the same time, this project demonstrated that for users familiar with Roulette and the LOP approach, Roulette provides powerful new capabilities that can be integrated with other Sandia modeling capabilities.

97 MATHEMATICS AND COMPUTING↗