Search NASA⌕ Search

SEARCH · Search NASA

Results for “binary uncertainty”

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 19 records

A Lagrangian dual method for two-stage robust optimization with binary uncertainties

This report presents a new exact method to calculate worst-case parameter realizations in two-stage robust optimization problems with categorical or binary-valued uncertain data. Traditional exact algorithms for these problems, notably Benders decomposition and column-and-constraint generation, compute worst-case parameter realizations by solving mixed-integer bilinear optimization subproblems. However, their numerical solution can be computationally expensive not only due to their resulting large size after reformulating the bilinear terms, but also because decision-independent bounds on their variables are typically unknown. We propose an alternative Lagrangian dual method that circumvents these difficulties and is readily integrated in either algorithm. We specialize the method to problems where the binary parameters switch on or off constraints as these are commonly encountered in applications, and discuss extensions to problems that lack relatively complete recourse and to those with integer recourse. Numerical experiments provide evidence of significant computational improvements over existing methods.

42 ENGINEERING↗

Exploring Explicit Uncertainty for Binary Analysis (EUBA)

Reverse engineering (RE) analysts struggle to address critical questions about the safety of binary code accurately and promptly, and their supporting program analysis tools are simply wrong sometimes. The analysis tools have to approximate in order to provide any information at all, but this means that they introduce uncertainty into their results. And those uncertainties chain from analysis to analysis. We hypothesize that exposing sources, impacts, and control of uncertainty to human binary analysts will allow the analysts to approach their hardest problems with high-powered analytic techniques that they know when to trust. Combining expertise in binary analysis algorithms, human cognition, uncertainty quantification, verification and validation, and visualization, we pursue research that should benefit binary software analysis efforts across the board. We find a strong analogy between RE and exploratory data analysis (EDA); we begin to characterize sources and types of uncertainty found in practice in RE (both in the process and in supporting analyses); we explore a domain-specific focus on uncertainty in pointer analysis, showing that more precise models do help analysts answer small information flow questions faster and more accurately; and we test a general population with domain-general sudoku problems, showing that adding "knobs" to an analysis does not significantly slow down performance. This document describes our explorations in uncertainty in binary analysis.

97 MATHEMATICS AND COMPUTING↗

Quantifying Epistemic Uncertainty in Binary Classification via Accuracy Gain

ABSTRACT Recently, a surge of interest has been given to quantifying epistemic uncertainty (EU), the reducible portion of uncertainty due to lack of data. We propose a novel EU estimator in the binary classification setting, as the posterior expected value of the empirical gain in accuracy between the current prediction and the optimal prediction. In order to validate the performance of our EU estimator, we introduce an experimental procedure where we take an existing dataset, remove a set of points, and compare the estimated EU with the observed change in accuracy. Through real and simulated data experiments, we demonstrate the effectiveness of our proposed EU estimator.

97 MATHEMATICS AND COMPUTING↗

Black hole spin in X-ray binaries: giving uncertainties an $\mathcal{f}$

The two established techniques for measuring black hole spin in X-ray binaries often yield conflicting results, which must be resolved before either method may be deemed robust. In practice, black hole spin measurements based on fitting the accretion disc continuum effectively do not marginalize over the colour-correction factor f col . This factor parametrizes spectral hardening of the disc continuum by the disc atmosphere, whose true properties are poorly constrained. We incorporate reasonable systematic uncertainties in f col into the eight (non-maximal) black hole spin measurements vetted by the disc continuum fitting community. In most cases, an f col uncertainty of ±0.2–0.3 dominates the black hole spin error budget. We go on to demonstrate that plausible departures in f col values from those adopted by the disc continuum fitting practitioners can bring the discrepant black hole spins into agreement with those from iron line modelling. Systematic uncertainties in f col , such as the effects of strong magnetization, should be better understood before dismissing their potentially dominant impact on the black hole spin error budget.

4U 1543–47↗

Binary pseudo-random array (BPRA) for inspection and calibration for cylindrical wavefront interferometry

High-accuracy metrology is vitally important in manufacturing ultra-high-quality free-form mirrors designed to manipulate X-ray light with nanometer-scale wavelengths. However, surface topography measurements are instrument dependent, and without the knowledge of how the instrument performs under the practical usage conditions, the measured data contain some degree of uncertainty. Binary Pseudo Random Array (BPRA) “white noise” artifact are effective and useful for characterizing the Instrument Transfer Function (ITF) of surface topography metrology tools and wavefront measurement instrument. BPRA artifact contains features with all spatial frequencies in the instrument bandpass with equal weight. As a result, power spectral density of the patterns has a deterministic white-noise-like character that allows direct determination of the ITF with uniform sensitivity over the entire spatial frequency range. The application examples include electron microscopes, x-ray microscopes, interferometric microscopes, and large field-of-view Fizeau Interferometers. Furthermore, we will introduce the application of BPRA method to characterizing the ITF of Cylindrical Wavefront Interferometry (CWI), by developing the BPRA artifact which matches the radius of curvature of the cylindrical wavefront. The data acquisition and analysis procedures for different applications of the ITF calibration technique developed are also discussed.

Munechika, K↗

Classification Using Support Vector Machines with Uncertainty Quantification

Binary classification using machine learning is needed to address engineering problems such as identifying passing/failing parts based on measured features from aging hardware. In these classifications, providing the uncertainty of each prediction is essential to support engineering decision making. One popular classifier is the support vector machine (SVM). There are many variations, with the simplest being a linear division between two classes with a hyperplane. Kernel methods can be implement

Taylor, Sofia Nitsche↗

BINARY CONTAMINATION IN THE SEGUE SAMPLE: EFFECTS ON SSPP DETERMINATIONS OF STELLAR ATMOSPHERIC PARAMETERS

We examine the effects that unresolved binaries have on the determination of various stellar atmospheric parameters for targets from the Sloan Extension for Galactic Understanding and Exploration (SEGUE) using numerical modeling, a grid of synthetic spectra, and the SEGUE Stellar Parameter Pipeline (SSPP). The SEGUE survey, a component of the Sloan Digital Sky Survey-II (SDSS-II) project focusing on Galactic structure, provides medium resolution spectroscopy for over 200,000 stars of various spectral types over a large area on the sky. To model undetected binaries that may be in this sample, we use a variety of mass distributions for the primary and secondary stars in conjunction with empirically determined relationships for orbital parameters to determine the fraction of G - K dwarf stars, defined by SDSS color cuts as having 0.48 {<=} (g - r){sub 0} {<=} 0.75, that will be blended with a secondary companion. We focus on the G-K dwarf sample in SEGUE as it records the history of chemical enrichment in our galaxy. To determine the effect of the secondary on the spectroscopic parameters, specifically effective temperature, surface gravity, metallicity, and [{alpha}/Fe], we synthesize a grid of model spectra from 3275 to 7850 K and [Fe/H] = -0.5 to -2.5 from MARCS model atmospheres using TurboSpectrum. These temperature and metallicity ranges roughly correspond to a stellar mass range of 0.1-1.0 M{sub sun}. We assume that both stars in the pair have the same metallicity. We analyze both 'infinite' signal-to-noise ratio (S/N) models and degraded versions of the spectra, at median S/N of 50, 25, and 10. By running individual and combined spectra (representing the binaries) through the SSPP, we determine that {approx}10% of the blended G - K dwarf pairs with S/N {>=} 25 will have their atmospheric parameter determinations, in particular temperature and metallicity, noticeably affected by the presence of an undetected secondary; namely, they will be shifted beyond the expected SSPP uncertainties. Shifts in [Fe/H] largely result from the shifts in temperature caused by a secondary. The additional uncertainty from binarity in targets with S/N {>=} 25 is {approx}80 K in temperature and {approx}0.1 dex in [Fe/H]. The effect on surface gravity and [{alpha}/Fe] is even smaller. As the S/N of targets decreases, the uncertainties from undetected secondaries increase. For S/N = 10, 40% of the G-K dwarf sample is shifted beyond expected uncertainties for this S/N in effective temperature and/or metallicity. To account for the additional uncertainty from binary contamination at an S/N {approx} 10, the most extreme scenario, uncertainties of {approx}140 K and {approx}0.17 dex in [Fe/H] must be added in quadrature to the published uncertainties of the SSPP.

79 ASTRONOMY AND ASTROPHYSICS↗

A DEEP RADIO SURVEY OF HARD STATE AND QUIESCENT BLACK HOLE X-RAY BINARIES

We have conducted a deep radio survey of a sample of black hole X-ray binaries in the hard and quiescent states to determine whether any systems were sufficiently bright for astrometric follow-up with high-sensitivity very long baseline interferometric arrays. The one hard-state system, Swift J1753.5-0127, was detected at a level of 0.5 mJy beam{sup -1}. All 11 quiescent systems were not detected. In the three cases with the highest predicted quiescent radio brightnesses (GRO J0422+32, XTE J1118+480, and GRO J1655-40), the new capabilities of the Expanded Very Large Array were used to reach noise levels as low as 2.6 {mu}Jy beam{sup -1}. None of the three sources were detected to 3{sigma} upper limits of 8.3, 7.8, and 14.2 {mu}Jy beam{sup -1}, respectively. These observations represent the most stringent constraints to date on quiescent radio emission from black hole X-ray binaries. The uncertainties in the source distances, quiescent X-ray luminosities at the times of the observations, and the power-law index of the empirical correlation between radio and X-ray luminosities make it impossible to determine whether these three sources are significantly less luminous in the radio band than expected. Thus it is not clear whether that correlation holds all the way down to quiescence for all black hole X-ray binaries.

79 ASTRONOMY AND ASTROPHYSICS↗

SDSS J092712.65+294344.0: RECOILING BLACK HOLE OR A SUBPARSEC BINARY CANDIDATE?

A search for recoiling supermassive black hole (SBH) candidates recently yielded the best candidate thus far, SDSS J092712.65+294344.0 reported by Komossa et al. Here we propose the alternative hypothesis that this object is an SBH binary. From the velocity shift imprinted in the emission-line spectrum we infer an orbital period of {approx}190 years for a binary mass ratio of 0.1, a secondary black hole mass of 10{sup 8} M {sub sun}, and assuming inclination and orbital phase angles of 45 deg. In this model the origin of the blueshifted narrow emission lines is naturally explained in the context of an accretion flow within the inner rim of the circumbinary disk. We attribute the blueshifted broad emission lines to gas associated with a disk around the accreting secondary black hole. We show that, within the uncertainties, this binary system can be long lived and thus, is not observed in a special moment in time. The orbital motion of the binary can potentially be observed with the VLBA if at least the secondary black hole is a radio emitter. In addition, for the parameters quoted above, the orbital motion will result in a {approx}100 km s{sup -1} velocity shift of the emission lines on a timescale of about a year, providing a direct observational test for the binary hypothesis.

79 ASTRONOMY AND ASTROPHYSICS↗

IN-SYNC. III. THE DYNAMICAL STATE OF IC 348—A SUPER-VIRIAL VELOCITY DISPERSION AND A PUZZLING SIGN OF CONVERGENCE

Most field stars will have encountered the highest stellar density and hence the largest number of interactions in their birth environment. Yet the stellar dynamics during this crucial phase are poorly understood. Here we analyze the radial velocities measured for 152 out of 380 observed stars in the 2–6 Myr old star cluster IC 348 as part of the SDSS-III APOGEE. The radial velocity distribution of these stars is fitted with one or two Gaussians, convolved with the measurement uncertainties including binary orbital motions. Including a second Gaussian improves the fit; the high-velocity outliers that are best fit by this second component may either (1) be contaminants from the nearby Perseus OB2 association, (2) be a halo of ejected or dispersing stars from IC 348, or (3) reflect that IC 348 has not relaxed to a Gaussian velocity distribution. We measure a velocity dispersion for IC 348 of 0.72 ± 0.07 km s{sup −1} (or 0.64 ± 0.08 km s{sup −1} if two Gaussians are fitted), which implies a supervirial state, unless the gas contributes more to the gravitational potential than expected. No evidence is found for a dependence of this velocity dispersion on distance from the cluster center or stellar mass. We also find that stars with lower extinction (in the front of the cloud) tend to be redshifted compared with stars with somewhat higher extinction (toward the back of the cloud). This data suggest that the stars in IC 348 are converging along the line of sight. We show that this correlation between radial velocity and extinction is unlikely to be spuriously caused by the small cluster rotation of 0.024 ± 0.013 km s{sup −1} arcmin{sup −1} or by correlations between the radial velocities of neighboring stars. This signature, if confirmed, will be the first detection of line of sight convergence in a star cluster. Possible scenarios for reconciling this convergence with IC 348's observed supervirial state include: (a) the cluster is fluctuating around a new virial equilibrium after a recent disruption due to gas expulsion or a merger event, or (b) the population we identify as IC 348 results from the chance alignment of two sub-clusters converging along the line of sight. Additional measurements of tangential and radial velocities in IC 348 will be important for clarifying the dynamics of this region and informing models of the formation and evolution of star clusters. The radial velocities analyzed in this paper have been made available online.

79 ASTRONOMY AND ASTROPHYSICS↗

Recent heavy-ion results with the ATLAS detector at the LHC

Results are presented from the ATLAS collaboration from the 2010 LHC heavy-ion run, during which nearly ten inverse microbarns of luminosity were delivered. Soft physics results include charged particle multiplicities and collective flow. The charged particle multiplicity, which tracks initial state entropy production, increases by a factor of 2 relative to the top RHIC energy, with a centrality dependence very similar to that already measured at the RHIC. Measurements of elliptic flow out to large transverse momentum also show similar results to what was measured at the RHIC, but no significant pseudorapidity dependence. Extensions of these measurements to higher harmonics have also been made and can be used to explain structures in the two-particle correlation functions that had long been attributed to jet-medium interactions. New hard probe measurements include single muons, jets and high p{sub T} hadrons. Single muons at high momentum are used to extract the yield of W{sup {+-}} bosons and are found to be consistent within statistical uncertainties with binary collision scaling. Conversely, jets are found to be suppressed in central events by a factor of 2 relative to peripheral events, with no significant dependence on the jet energy. Fragmentation functions are also found to be the same in central and peripheral events. Finally, charged hadrons have been measured out to 30 GeV, and their centrality dependence relative to peripheral events is similar to that found for jets.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Conditional Tree Reduction in the ADAPT Framework

Probabilistic Risk Assessment (PRA) uses fault tree/event tree analysis to evaluate the environmental impacts of nuclear power plants due to internal or external initiating events. In a fault tree, basic failure events are assembled using primarily And/Or logic to determine the combinations of failures that lead to failure of the system as a whole. An event tree is forward-facing, and begins with a single initiating event. From there the analysis considers what event may occur next, and branches out among the possible configurations of occurrence and non-occurrence of that event. This branching continues until end states are reached. End states may include a safe and stable configuration of the plant, or any number of differing failure states. Both fault and event trees require basic event probabilities as inputs to provide insight on the likelihood of different outcomes. PRA has been applied to nuclear power plants as an analysis tool since the 1975 Reactor Safety Study. In a traditional PRA, the order of events is prescribed by the analyst and each event is typically a binary aleatory uncertainty: occurrence or non-occurrence. Discrete dynamic event tree (DDET) analysis eliminates this subjectivity in ordering of events by using the output of a dynamic system model (simulator) to inform the branching. Branching conditions are triggered by the existence of a relevant plant state in the code, and therefore only occur as physically appropriate. The DDET approach also allows consideration of both epistemic and aleatory uncertainties on a phenomenologically and stochastically consistent platform. A DDET may be orders of magnitude larger than a traditional event tree for the same initiating event, as more of the uncertainty space is likely to be explored and the tree does not require manual assembly. Because of the size, manual inspection of the entire DDET is often infeasible. This problem may be approached from the front by reducing the number of sampled branching values or by online pruning of the tree, thus reducing the total number of branches. Utilizing post-processing techniques to draw actionable insights via clustering has also been proposed. One post-processing tool to reduce the DDET according to user-defined rules has been recently implemented in the Analysis of Dynamic Accident Progression Trees (ADAPT) DDET driver code. The tool creates a 'slice' of the DDET comprised of the set of branches that meet a set of user-input rules, for immediate visualization or for further calculations conditional on the rules being met. This extension is expected to ease analysis of the results of large and complex DDETs, leading to greater applicability of DDETs in PRA. It should be emphasized that this tool is a visualization aide and does not impact the theoretical basis of the DDET approach. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Extension of the ADAPT Framework for Multiple Simulators

Probabilistic Risk Assessment (PRA) uses fault tree/event tree analysis to evaluate the environmental impacts of nuclear power plants due to internal or external initiating events. In a fault tree, basic failure events are assembled using primarily And/Or logic to determine the combinations of failures that lead to failure of the system as a whole. An event tree is forward-facing, and begins with a single initiating event. From there the analysis considers what event may occur next, and branches out among the possible configurations of occurrence and non-occurrence of that event. This branching continues until end states are reached. End states may include a safe and stable configuration of the plant, or any number of differing failure states. Both fault and event trees require basic event probabilities as inputs to provide insight on the likelihood of different outcomes. PRA has been applied to nuclear power plants as an analysis tool since the 1975 Reactor Safety Study. In a traditional PRA, the order of events is prescribed by the analyst and each event is typically a binary aleatory uncertainty: occurrence or non-occurrence. Discrete dynamic event tree (DDET) analysis eliminates this subjectivity in ordering of events by using the output of a dynamic system model (simulator) to inform the branching. Branching conditions are triggered by the existence of a relevant plant state in the code, and therefore only occur as physically appropriate. The DDET approach also allows consideration of both epistemic and aleatory uncertainties on a phenomenologically and stochastically consistent platform. One limitation of the DDET driver codes developed to date (namely ADAPT, ADS, and MCDET ) is that the tree is driven by a single simulator. The goal of this work is a general multi-simulator driver. This limits the parameter space that may be explored in a single DDET to that which may be covered by a single simulator. In this work, the Analysis of Dynamic Accident Progression Trees (ADAPT) code is extended to support multiple arbitrary simulators. This capability will increase the depth and breadth of phenomena that may be analyzed in a single DDET. For example, a large-scale system code may pass the analysis to a more detailed code if certain conditions are met, in order to capture relevant phenomena that would be lost if only the system code were to be used. In addition, the time required to cover a given uncertainty space may be reduced by dynamically identifying and running only the situations that are appropriate given the plant conditions in each simulator. The use of multiple simulators is expected to further reduce subjectivity in PRA by enabling high fidelity modeling of reality. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗