Search NASA⌕ Search

SEARCH · Search NASA

Results for “random sampling”

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 217 records · Page 12

A Statistical Framework for Microbial Source Attribution

This report presents a general approach to inferring transmission and source relationships among microbial isolates from their genetic sequences. The outbreak transmission graph (also called the transmission tree or transmission network) is the fundamental structure which determines the statistical distributions relevant to source attribution. The nodes of this graph are infected individuals or aggregated sub-populations of individuals in which transmitted bacteria or viruses undergo clonal expansion, leading to a genetically heterogeneous population. Each edge of the graph represents a transmission event in which one or a small number of bacteria or virions infects another node thus increasing the size of the transmission network. Recombination and re-assortment events originate in nodes which are common to two distinct networks. In order to calculate the probability that one node was infected by another, given the observed genetic sequences of microbial isolates sampled from them, we require two fundamental probability distributions. The first is the probability of obtaining the observed mutational differences between two isolates given that they are separated by M steps in a transmission network. The second is the probability that two nodes sampled randomly from an outbreak transmission network are separated by M transmission events. We show how these distributions can be obtained from the genetic sequences of isolates obtained by sampling from past outbreaks combined with data from contact tracing studies. Realistic examples are drawn from the SARS outbreak of 2003, the FMDV outbreak in Great Britain in 2001, and HIV transmission cases. The likelihood estimators derived in this report, and the underlying probability distribution functions required to calculate them possess certain compelling general properties in the context of microbial forensics. These include the ability to quantify the significance of a sequence 'match' or 'mismatch' between two isolates; the ability to capture non-intuitive effects of network structure on inferential power, including the 'small world' effect; the insensitivity of inferences to uncertainties in the underlying distributions; and the concept of rescaling, i.e. ability to collapse sub-networks into single nodes and examine transmission inferences on the rescaled network.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Uncertainty Estimates for Fitting Zernike Polynomials to Discrete Data

Zernike polynomials are a widely used metric in modern optical analysis. They conveniently represent surfaces as a series of weighted terms corresponding to various optical aberrations. Ideally, each term is independent of others in the series, but Zernike polynomials lose this property when working with sets of discrete data. This gives rise to uncertainty in each polynomial’s actual contribution and affects metrology and simulation estimates of their relative weights. Several factors influencing these estimates are the number and arrangement of sample locations, the method for calculating the weights, and the total number of Zernike terms used in the calculation. Discussed is the uncertainty associated with linear regression using random sampling. Other topics reviewed are complex Zernike polynomials and vector spaces of functions.

Zernike Polynomials↗

Stochastic sampling method with MCNPX for nuclear data uncertainty propagation in criticality safety applications

In the domain of criticality safety, the efficient propagation of uncertainty in nuclear data to uncertainty in k{sub eff} is an important area of current research. In this paper, a method based on stochastic sampling is presented for uncertainty propagation in MCNPX calculations. To that aim, the nuclear data (i.e. cross sections) are assumed to have a multivariate normal distribution and simple random sampling is performed following this presumed probability distribution. A verification of the developed stochastic sampling procedure with MCNPX is then conducted using the {sup 239}Pu Jezebel experiment as well as the PB-2 BWR and TMI-1 PWR pin cell models from the Uncertainty Analysis in Modeling (UAM) exercises. For the Jezebel case, it is found that the developed stochastic sampling approach predicts similar k{sub eff} uncertainties compared to conventional sensitivity and uncertainty methods. For the UAM models, slightly lower uncertainties are obtained when comparing to existing preliminary results. Further details of these verification studies are discussed and directions for future work are outlined. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

A Case Study in Assessing a Potential Severity Framework for Incidents from a Decadal Sample

In this study, the primary objective of this case study is to determine the applicability and feasibility of a framework that leverages occupational incident details to prospectively identify “potential Serious Injury or Fatality” (pSIF) cases. This study comprehensively reviewed a random sample of 1,081 injury and illness cases across 21 generalized incident types spanning over a decade at Lawrence Livermore National Laboratory (LLNL), a U.S. Department of Energy research and development facility with more than 9,000 employees. The review applied a general framework that classified each case on information suitability, potential severity, and future incident mitigation. The findings from the study indicate that 86.6% of the cases had sufficient information to make a high-confidence determination on potential severity, underscoring the feasibility of applying this general framework. Additionally, cases with a higher pSIF score had, on average, a higher level of institutional response. Implementing a simplified methodology for incident classification that emphasizes incidents that pose high potential severity, regardless of incident type, can help LLNL prioritize resources and tailor responses to such incidents using a graded approach. LLNL has recognized the value of this capability and is integrating the framework into their injury and illness process in the 2024 calendar year.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗

Computing the QRPA level density with the finite amplitude method

Here, we describe a new algorithm to calculate the vibrational nuclear level density of an atomic nucleus. Fictitious perturbation operators that probe the response of the system are generated by drawing their matrix elements from some probability distribution function. We use the Finite Amplitude Method to explicitly compute the response for each such sample. With the help of the Kernel Polynomial Method, we build an estimator of the vibrational level density and provide the upper bound of the relative error in the limit of infinitely many random samples. The new algorithm can give accurate estimates of the vibrational level density. Since it is based on drawing multiple samples of perturbation operators, its computational implementation is naturally parallel and scales like the number of available processing units.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Assessing the Uncertainty in QUANTEC's Dose–Response Relation of Lung and Spinal Cord With a Bootstrap Analysis

Purpose: To apply a statistical bootstrap analysis to assess the uncertainty in the dose–response relation for the endpoints pneumonitis and myelopathy reported in the QUANTEC review. Methods and Materials: The bootstrap method assesses the uncertainty of the estimated population-based dose-response relation due to sample variability, which reflects the uncertainty due to limited numbers of patients in the studies. A large number of bootstrap replicates of the original incidence data were produced by random sampling with replacement. The analysis requires only the dose, the number of patients, and the number of occurrences of the studied endpoint, for each study. Two dose–response models, a Poisson-based model and the Lyman model, were fitted to each bootstrap replicate using maximum likelihood. Results: The bootstrap analysis generates a family of curves representing the range of plausible dose–response relations, and the 95% bootstrap confidence intervals give an estimated upper and lower toxicity risk. The curve families for the 2 dose–response models overlap for doses included in the studies at hand but diverge beyond that, with the Lyman model suggesting a steeper slope. The resulting distributions of the model parameters indicate correlation and non-Gaussian distribution. For both data sets, the likelihood of the observed data was higher for the Lyman model in >90% of the bootstrap replicates. Conclusions: The bootstrap method provides a statistical analysis of the uncertainty in the estimated dose–response relation for myelopathy and pneumonitis. It suggests likely values of model parameter values, their confidence intervals, and how they interrelate for each model. Finally, it can be used to evaluate to what extent data supports one model over another. For both data sets considered here, the Lyman model was preferred over the Poisson-based model.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

hiposa v0.1.0

Fast construction of hierarchical N-dimensional experimental design. The software allows construction of sampling strategies that are self-avoiding, tunable in resolution in N dimensions. This code is a drop-in replacement for seeding optimization runs often performed using uniform random sampling. The hierarchical approach allows for a rapid, top down method for construction the most effective design that drives towards a global optimum.

Zwart, PetrusH [Lawrence Berkeley National Laborat↗

In-service inspection of radioactive waste tanks at the Savannah River Site - 15410

Liquid radioactive wastes from the Savannah River Site (SRS) separation process are stored in large underground carbon steel tanks. The high level wastes are processed in several of the tanks and then transferred by piping to other site facilities for further processing before they are stabilized in a vitrified or grout waste form. Based on waste removal and processing schedules, many of the tanks, will be required to be in service for times exceeding the initial intended life. Until the waste is removed from storage, transferred, and processed, the materials and structures of the tanks must maintain a confinement function by providing a barrier to the environment and by maintaining acceptable structural stability during design basis events, which include loadings from both normal service and abnormal (e.g., earthquake) conditions. A structural integrity program is in place to maintain the structural and leak integrity functions of these waste tanks throughout their intended service life. In-service inspection (ISI) is an essential element of a comprehensive structural integrity program for the waste tanks at the Savannah River Site (SRS). The ISI program was developed to determine the degree of degradation the waste tanks have experienced due to service conditions. As a result of the inspections, an assessment can be made of the effectiveness of corrosion controls for the waste chemistry, which precludes accelerated localized and general corrosion of the waste tanks. Ultrasonic inspections (UT) are performed to detect and quantify the degree of general wall thinning, pitting and cracking as a measure of tank degradation. The results from these inspections through 2013, for the 27 Type III/IIIA tanks, indicate no reportable in-service corrosion degradation in the primary tank (i.e., general, pitting, or cracking). The average wall thickness for all tanks remains above the manufactured nominal thickness minus 0.25 millimeter and the largest pit identified is approximately 1.70 millimeter deep (i.e., less than 10% through-wall). Improvements to the inspection program were recently instituted to provide additional confidence in the degradation rates. Thickness measurements from a single vertical strip along the accessible height of the primary tank have been used as a baseline to compare historical measurements. Changes in wall thickness and pit depths along this vertical strip are utilized to estimate the rate of corrosion degradation. An independent review of the ISI program methodology, results, and path forward was held in August 2009. The review recommended statistical sampling of the tanks to improve the confidence of the single strip inspection program. The statistical sampling plan required that SRS increase the amount of area scanned per tank. Therefore, in addition to the baseline vertical strip that is obtained for historical comparisons, four additional randomly selected vertical strips are inspected. To date, a total of 104 independent vertical strips along the height of the primary tank have been completed. A statistical analysis of the data indicates that at this coverage level there is a 99.5% confidence level that one of the worst 5% of all the vertical strips was inspected. That is, there is a relatively high likelihood that the SRS inspection program has covered one of the most corroded areas of any of the Type III/IIIA waste tanks. These data further support the conclusion that there are no significant indications of wall thinning or pitting. Random sampling will continue to increase the confidence that one of the worst 5% has been inspected. In order to obtain the additional vertical strips, and minimize budget and schedule impacts, data collection speed for the UT system was optimized. Prior to 2009, the system collected data at a rate of 32 square centimeters per minute. The scan rate was increased to 129 - 160 square centimeters per minute by increasing the scanner step and pixel sizes in the data acquisition set-up. Laboratory testing was utilized to optimize the scan index/pixel size such that the requirements for wall thinning and pit detection were still maintained. SRS continues to evaluate improvements to ultrasonic equipment. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Energy Delivery Systems with Verifiable Trustworthiness (Final Report)

Energy Delivery Systems (EDS) must be verified to be free from intrusive and malicious software. One way of verifying this software is to perform device scans to detect malicious code. Because it is possible to have “fileless” malware that exists only in device (volatile) memory, offline scanning and even many forms of online scanning is insufficient for detection. This project (“Verify”) addresses this need by performing direct sampling of memory during device operation to detect unexpected or modified software while not interfering with device operation. The Verify project provides a proof-of-concept of detection by random sampling combined with remote software- and timing-based attestation methods for robust detection of in-memory threats. An external review of Verify was performed by our partner, General Electric (GE), and a summary of their findings is provided.

97 MATHEMATICS AND COMPUTING↗

High-Dimensional Spectral Sampling

A wide variety of applications in science and engineering depend on effectively and efficiently exploring high- dimensional parameter spaces. Common examples are uncertainty quantification, inverse problems, or various optimization approaches. In such applications, the first step is to create an initial uniform, random sampling of the space to create a baseline of knowledge. The simulations evaluated at these samples are then used for subsequent processing. In its most generic form, the goal of sampling is to produce the maximal amount of information with the minimal number of samples. However, despite being the first step and thus influencing the performance of the entire workflow, the quality of the initial sample designs is rarely analyzed or optimized. This project develops a new spectral sampling theory for analyzing and creating experiment designs in high-dimensional spaces. Further, new algorithms for estimating sample properties and sample synthesis are designed. Finally, applications of the proposed spectral sampling tools in surrogate modeling, scientific machine learning and image analysis are presented.

97 MATHEMATICS AND COMPUTING↗

A Machine Learning-Based Vulnerability Analysis for Cascading Failures of Integrated Power-Gas Systems

This article proposes a cascading failure simulation (CFS) method and a hybrid machine learning method for vulnerability analysis of integrated power-gas systems (IPGSs). The CFS method is designed to study the propagating process of cascading failures between the two systems, generating data for machine learning with initial states randomly sampled. The proposed method considers generator and gas well ramping, transmission line and gas pipeline tripping, island issue handling and load shedding strategies. Then, a hybrid machine learning model with a combined random forest (RF) classification and regression algorithms is proposed to investigate the impact of random initial states on the vulnerability metrics of IPGSs. Extensive case studies are carried out on three test IPGSs to verify the proposed models and algorithms. Simulation results show that the proposed models and algorithms can achieve high accuracy for the vulnerability analysis of IPGSs.

24 POWER TRANSMISSION AND DISTRIBUTION↗

A Large Ordinary Chondrite Shower in the Dominion Range

The US Antarctic Meteorite Program has visited the Dominion Range in the Transantarctic Mountains during several different seasons, including the 1985, 2003, 2008, 2010, and 2014 seasons. Total recovered meteorites from this region is over 2000. The 2008 and 2010 seasons have been fully classified and, respectively) revealing the presence of a large meteorite shower that comprises approximately 60% of all samples recovered in those two seasons. The oil immersion classification suggests that this shower is LL chondrite material, whereas published magnetic susceptibility (MS; log chi) measurements yield L chondrite values. However, usually random sampling of a large collection like this would uncover EOC material for which we have prepared thin sections. In this case, no LL chondrite materials have been found in thin section, suggesting that the shower might instead be an L chondrite. L and LL chondrites are notoriously difficult to distinguish using oil immersion techniques. To better characterize this large group of samples, we have decided to examine some of the large members of this group, using EMPA analysis of the olivines to verify the classifications. With a compositional link between this subset of samples, and the MS measurements, we can more confidently classify the samples making up this pairing group. Subsequently, more accurate and meaningful comparisons may be drawn between this pairing group and some other Antarctic pairing groups such as from the Queen Alexandra Range (QUE), and Lewis Cliffs Ice Tongue (LEW). electron microprobe analysis

Satterwhite, C. E.↗

Modeling the determinants of the social impacts of agricultural development projects

In an attempt to help policy-makers improve the social sustainability of development projects, this study identifies the key determinants of farmers' attitudes relating to the social impacts of the floodwater spreading project (FWSP) on the Gareh-Bygone plain in Iran. In order to analyze the links between the various factors that affect the experience of social impact, a theoretical framework was developed. Stratified random sampling was used to survey 138 farm households from the four villages in the region. One male and one female from each house were interviewed face-to-face using a questionnaire, resulting in a total of 276 interviews. Structural factors were found to be the largest contribution to stakeholders' attitudes relating to the social impacts of the project. Results from a cluster analysis suggested that the level of floodwater information, level of participation, water access, ownership change, and environmental worldview were the most important factors explaining attitude towards social impact of the FWSP.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

Survey of Antihomology Research

Homology has been gaining traction as a tool for studying patterns in real-world datasets. During the past year, I have sought to understand a slight variant on the typical computation of homology of a metric space which we term antihomology . This exposition will survey that work. In section 2, I begin with findings from computations on samples from common spaces (the circle, the disc, and the square). Section 3 presents one conjecture and one lemma regarding 1-dimensional cycles on the circle with evenly space points and the circle with random sampling. In section 4, I applied antihomology to power grid datasets and compared results with more traditional data reduction techniques. Lastly, section 5 reveals the connection between antihomology and location covering problems.

97 MATHEMATICS AND COMPUTING↗

SUNYAEV-ZEL'DOVICH EFFECT OBSERVATIONS OF STRONG LENSING GALAXY CLUSTERS: PROBING THE OVERCONCENTRATION PROBLEM

We have measured the Sunyaev-Zel'dovich (SZ) effect for a sample of 10 strong lensing selected galaxy clusters using the Sunyaev-Zel'dovich Array (SZA). The SZA is sensitive to structures on spatial scales of a few arcminutes, while the strong lensing mass modeling constrains the mass at small scales (typically <30''). Combining the two provides information about the projected concentrations of the strong lensing clusters. The Einstein radii we measure are twice as large as expected given the masses inferred from SZ scaling relations. A Monte Carlo simulation indicates that a sample randomly drawn from the expected distribution would have a larger median Einstein radius than the observed clusters about 3% of the time. The implied overconcentration has been noted in previous studies and persists for this sample, even when we take into account that we are selecting large Einstein radius systems, suggesting that the theoretical models still do not fully describe the observed properties of strong lensing clusters.

79 ASTRONOMY AND ASTROPHYSICS↗

Understanding the Awareness, Knowledge and Perceptions of Interventional Radiology Amongst Undergraduates in the UK

Introduction: The role of interventional radiology (IR) in health care is rapidly expanding in the United Kingdom (UK). However, the presence of IR in the current undergraduate curriculum is lacking. Purpose: This study investigates the awareness, knowledge and perceptions of IR in undergraduates through the clinical years of medical school to suggest future interventions. Methods: A total of 360 questionnaires, designed to test undergraduates’ knowledge of, and exposure to, IR were distributed over a 14-day period in a UK medical institution. Results: Of the 312 respondents (86.7%), the majority had never formally shadowed a radiologist (65.4%) or encountered teaching from an interventional radiologist (54.5%). Most students (88.5%) felt their knowledge of IR was lower than other specialties. Students’ knowledge of IR procedures did not statistically increase through the clinical years. Students were most cognisant of vascular procedures and least aware of IR’s role in oncology. Only 10.9% would consider a career in IR, with 39.7% citing the lack of knowledge and 18.6% citing patient contact as the top reasons for this. Conclusion: Current students have poor exposure to IR through the undergraduate curriculum. Consequently, there is no increase in knowledge of IR clinical practice through the clinical years. Multi-level interventions, incorporating universities and radiological societies are required to increase undergraduate awareness and promote IR as a career. Level of Evidence: Level 3, Local non-random sample.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Improved statistical models for limited datasets in uncertainty quantification using stochastic collocation

This paper presents a data-driven framework for performing uncertainty quantification (UQ) by choosing a stochastic model that accurately describes the sources of uncertainty in a system. This model is propagated through an appropriate response surface function that approximates the behavior of this system using stochastic collocation. Given a sample of data describing the uncertainty in the inputs, our goal is to estimate a probability density function (PDF) using the kernel moment matching (KMM) method so that this PDF can be used to accurately reproduce statistics like mean and variance of the response surface function. Instead of constraining the PDF to be optimal for a particular response function, we show that we can use the properties of stochastic collocation to make the estimated PDF optimal for a wide variety of response functions. We contrast this method with other traditional procedures that rely on the Maximum Likelihood approach, like kernel density estimation (KDE) and its adaptive modification (AKDE). We argue that this modified KMM method tries to preserve what is known from the given data and is the better approach when the available data is limited in quantity. We test the performance of these methods for both univariate and multivariate density estimation by sampling random datasets from known PDFs and then measuring the accuracy of the estimated PDFs, using the known PDF as a reference. Comparing the output mean and variance estimated with the empirical moments using the raw data sample as well as the actual moments using the known PDF, we show that the KMM method performs better than KDE and AKDE in predicting these moments with greater accuracy. This improvement in accuracy is also demonstrated for the case of UQ in electrostatic and electrothermomechanical microactuators. We show how our framework results in the accurate computation of statistics in micromechanical systems.

97 MATHEMATICS AND COMPUTING↗