Search NASA⌕ Search

SEARCH · Search NASA

Results for “Randomized methods”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,081 records · Page 60

Fabrication of oriented crystals as force measurement tips via focused ion beam and microlithography methods

Detailed knowledge of the forces between nanocrystals is very crucial for understanding many generic (e.g., random aggregation/assembly and rheology) and specific (e.g., oriented attachment) phenomena at macroscopic length scales, especially considering the additional complexities involved in nanocrystals such as crystal orientation and corresponding orientation-dependent physicochemical properties. Because there are a limited number of methods to directly measure the forces, little is known about the forces that drive the various emergent phenomena. Here we report on two methods of preparing crystals as force measurement tips used in an atomic force microscope (AFM): the focused ion beam method and microlithography method. The desired crystals are fabricated using these two methods and are fixed to the AFM probe using platinum deposition, ultraviolet epoxy, or resin, which allows for the orientation-dependent force measurements. These two methods can be used to attach virtually any solid particles (from the size of a few hundreds of nanometers to millimeters). We demonstrate the force measurements between aqueous media under different conditions such as pH.

Environmental Molecular Sciences Laboratory↗

Vehicle track segmentation using higher order random fields

Here, we present an approach to segment vehicle tracks in coherent change detection images, a product of combining two synthetic aperture radar images taken at different times. The approach uses multiscale higher order random field models to capture track statistics, such as curvatures and their parallel nature, that are not currently utilized in existing methods. These statistics are encoded as 3-by-3 patterns at different scales. The model can complete disconnected tracks often caused by sensor noise and various environmental effects. Coupling the model with a simple classifier, our approach is effective at segmenting salient tracks. We improve the F-measure on a standard vehicle track data set to 0.963, up from 0.897 obtained by the current state-of-the-art method.

97 MATHEMATICS AND COMPUTING↗

An analytical model of the large neutral regions during the late stage of reionization

In this paper, we investigate the nature and distribution of large neutral regions during the late epoch of reionization. In the 'bubble model' of reionization, the mass distribution of large ionized regions ('bubbles') during the early stage of reionization is obtained by using the excursion set model, where the ionization of a region corresponds to the first up-crossing of a barrier by random trajectories. We generalize this idea and develop a method to predict the distribution of large-scale neutral regions during the late stage of reionization, taking into account the ionizing background after the percolation of H II regions. The large-scale neutral regions, which we call 'neutral islands', are not individual galaxies or minihalos, but larger regions where fewer galaxies formed and hence ionized later and they are identified in the excursion set model with the first down-crossings of the island barrier. Assuming that the consumption rate of ionizing background photons is proportional to the surface area of the neutral islands, we obtained the size distribution of the neutral islands. We also take the 'bubbles-in-island' effect into account by considering the conditional probability of up-crossing a bubble barrier after down-crossing the island barrier. We find that this effect is very important. An additional barrier is set to avoid islands being percolated through. We find that there is a characteristic scale for the neutral islands, while the small islands are rapidly swallowed up by the ionizing background; this characteristic scale does not change much as the reionization proceeds.

79 ASTRONOMY AND ASTROPHYSICS↗

Improved renormalization group computation of likelihood functions for cosmological data sets

Evaluation of likelihood functions for cosmological large scale structure data sets (including CMB, galaxy redshift surveys, etc.) naturally involves marginalization, i.e., integration, over an unknown underlying random signal field. Recently, I showed how a renormalization group method can be used to carry out this integration efficiently by first integrating out the smallest scale structure, i.e., localized structure on the scale of differences between nearby data cells, then combining adjacent cells in a coarse graining step, then repeating this process over and over until all scales have been integrated. Here I extend the formulation in several ways in order to reduce the prefactor on the method's linear scaling with data set size. The key improvement is showing how to integrate out the difference between specific adjacent cells before summing them in the coarse graining step, compared to the original formulation in which small-scale fluctuations were integrated more generally. I suggest some other improvements in details of the scheme, including showing how to perform the integration around a maximum likelihood estimate for the underlying random field. In the end, an accurate likelihood computation for a million-cell Gaussian test data set runs in two minutes on my laptop, with room for further optimization and straightforward parallelization.

79 ASTRONOMY AND ASTROPHYSICS↗

Randomized, Double-Blinded, Placebo-Controlled, Trial of Risedronate for the Prevention of Bone Mineral Density Loss in Nonmetastatic Prostate Cancer Patients Receiving Radiation Therapy Plus Androgen Deprivation Therapy

Purpose: Androgen deprivation therapy (ADT) has been used as an adjuvant treatment to radiation therapy (RT) for the management of locally advanced prostate carcinoma. Long-term ADT decreases bone mineral density (BMD) and increases the risk of osteoporosis. The objective of this clinical trial was to evaluate the efficacy of risedronate for the prevention of BMD loss in nonmetastatic prostate cancer patients undergoing RT plus 2 to 3 years of ADT. Methods and Materials: A double-blinded, placebo-controlled, randomized trial was conducted for nonmetastatic prostate cancer patients receiving RT plus 2 to 3 years of ADT. All had T scores > −2.5 on dual energy x-ray absorptiometry at baseline. Patients were randomized 1:1 between risedronate and placebo for 2 years. The primary endpoints were the percent changes in the BMD of the lumbar spine at 1 and 2 years from baseline, measured by dual energy x-ray absorptiometry. Analyses of the changes in BMD and bone turnover biomarkers were carried out by comparing mean values of the intrapatient changes between the 2 arms, using standard t tests. Results: One hundred four patients were accrued between 2004 and 2007, with 52 in each arm. Mean age was 66.8 and 67.5 years for the placebo and risedronate, respectively. At 1 and 2 years, mean (±SE) BMD of the lumbar spine decreased by 5.77% ± 4.66% and 13.55% ± 6.33%, respectively, in the placebo, compared with 0.12% ± 1.29% at 1 year (P=.2485) and 0.85% ± 1.56% (P=.0583) at 2 years in the risedronate. The placebo had a significant increase in serum bone turnover biomarkers compared with the risedronate. Conclusions: Weekly oral risedronate prevented BMD loss at 2 years and resulted in significant suppression of bone turnover biomarkers for 24 months for patients receiving RT plus 2 to 3 years of ADT.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Statistical uncertainty of eddy covariance CO 2 fluxes inferred using a residual bootstrap approach

High-frequency eddy-covariance measurements of net ecosystem CO 2 exchange (NEE) with the atmosphere are valuable resources for model parameterization, calibration, and validation. However, uncertainties in measured data, i.e., data gaps and inherent random errors, create problems for researchers attempting to quantify uncertainties in model projections of terrestrial ecosystem carbon cycling. Here, we demonstrate that a model-data fusion method (residual bootstrap) produces defensible annual NEE sums, through mimicking the behavior of random errors, filling missing values, and simulating gap-filling biases. This study estimated annual NEE sums for 53 site-years based on nine eddy-covariance tower sites in the USA, and found that our annual estimates were, in most cases, comparable in magnitude with those obtained from AmeriFlux gap-filled data. Additionally, compared to the AmeriFlux standardized gap-filling, our approach provides better NEE estimates for moderate to longer, and more frequent, data gaps. Annual accumulated uncertainties in NEE at the 95% confidence level were ±30gCm -2 year -1 for evergreen needleleaf forests; ±60gCm -2 year -1 for deciduous broadleaf forests; and ±80gCm -2 year -1 for croplands. The residual bootstrap performed worst when gap length was greater than one month or data exclusion greater than 90% during the growing season, common to other gap-filling techniques. However, this study produced robust results for most site years when monthly data coverage during the growing season is not extremely low. We therefore suggest that the inclusion of NEE uncertainty estimates and better estimation for moderate to longer, and more frequent, data gaps as provided by the residual bootstrap approach can be beneficial for ecosystem model evaluation.

54 ENVIRONMENTAL SCIENCES↗

Feasibility study of noise analysis methods on virtual thermal reactor subcriticality monitoring

This paper presents the analysis results of Rossi-alpha, cross-correlation, Feynman-alpha, and Feynman difference methods applied to the subcriticality monitoring of nuclear reactors. A thermal spectrum Godiva model has been designed for the analysis of the four methods. This Godiva geometry consists of a spherical core containing the isotopes of H-l, U-235 and U-238, and the H{sub 2}O reflector outside the core. A Monte Carlo code, McCARD, is used in real time mode to generate virtual detector signals to analyze the feasibility of the four methods. The analysis results indicate that the four methods can be used with high accuracy for the continuous monitoring of subcriticality. In addition to that, in order to analyze the impact of the random noise contamination on the accuracy of the noise analysis, the McCARD-generated signals are contaminated with arbitrary noise. It is noticed that, even when the detector signals are contaminated, the four methods can predict the subcriticality with reasonable accuracy. Nonetheless, in order to reduce the adverse impact of the random noise, eight detector signals, rather than a single signal, are generated from the core, one signal from each equally divided eighth part of the core. The preliminary analysis with multiple virtual detector signals indicates that the approach of using many detectors is promising to improve the accuracy of criticality prediction and further study will be performed in this regard. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Accelerating Random Forest Classification on GPU and FPGA

Random Forests (RFs) are a commonly used machine learning method for classification and regression tasks spanning a variety of application domains, including bioinformatics, business analytics, and software optimization. While prior work has focused primarily on improving performance of the training of RFs, many applications, such as malware identification, cancer prediction, and banking fraud detection, require fast RF classification. In this work, we accelerate RF classification on GPU and FPGA. In order to provide efficient support for large datasets, we propose a hierarchical memory layout suitable to the GPU/FPGA memory hierarchy. We design three RF classification code variants based on that layout, and we investigate GPU- and FPGA-specific considerations for these kernels. Our experimental evaluation, performed on an Nvidia Xp GPU and on a Xilinx Alveo U250 FPGA accelerator card using publicly available datasets on the scale of millions of samples and tens of features, covers various aspects. First, we evaluate the performance benefits of our hierarchical data structure over the standard compressed sparse row (CSR) format. Second, we compare our GPU implementation with cuML, a machine learning library targeting Nvidia GPUs. Third, we explore the performance/accuracy tradeoff resulting from the use of different tree depths in the RF. Finally, we perform a comparative performance analysis of our GPU and FPGA implementations. Our evaluation shows that for high accuracy targets, our GPU implementation yields 5-9x speedup over CSR, and up to a 2x speedup over cuML.

FPGA, Xilinx FPGA, GPU, Random Forest classificati↗

Machine-learning error models for approximate solutions to parameterized systems of nonlinear equations

Here, this work proposes a machine-learning framework for constructing statistical models of errors incurred by approximate solutions to parameterized systems of nonlinear equations. These approximate solutions may arise from early termination of an iterative method, a lower-fidelity model, or a projection-based reduced-order model, for example. The proposed statistical model comprises the sum of a deterministic regression-function model and a stochastic noise model. The method constructs the regression-function model by applying regression techniques from machine learning (e.g., support vector regression, artificial neural networks) to map features (i.e., error indicators such as sampled elements of the residual) to a prediction of the approximate-solution error. The method constructs the noise model as a mean-zero Gaussian random variable whose variance is computed as the sample variance of the approximate-solution error on a test set; this variance can be interpreted as the epistemic uncertainty introduced by the approximate solution. This work considers a wide range of feature-engineering methods, data-set-construction techniques, and regression techniques that aim to ensure that (1) the features are cheaply computable, (2) the noise model exhibits low variance (i.e., low epistemic uncertainty introduced), and (3) the regression model generalizes to independent test data. Finally, numerical experiments performed on several computational-mechanics problems and types of approximate solutions demonstrate the ability of the method to generate statistical models of the error that satisfy these criteria and significantly outperform more commonly adopted approaches for error modeling.

97 MATHEMATICS AND COMPUTING↗

Stochastic gradient descent algorithm for stochastic optimization in solving analytic continuation problems

We propose a stochastic gradient descent based optimization algorithm to solve the analytic continuation problem in which we extract real frequency spectra from imaginary time Quantum Monte Carlo data. The procedure of analytic continuation is an ill-posed inverse problem which is usually solved by regularized optimization methods, such like the Maximum Entropy method, or stochastic optimization methods. The main contribution of this work is to improve the performance of stochastic optimization approaches by introducing a supervised stochastic gradient descent algorithm to solve a flipped inverse system which processes the random solutions obtained by a type of Fast and Efficient Stochastic Optimization Method.

97 MATHEMATICS AND COMPUTING↗

Method and apparatus for routing data in an inter-nodal communications lattice of a massively parallel computer system by semi-randomly varying routing policies for different packets

A massively parallel computer system contains an inter-nodal communications network of node-to-node links. Nodes vary a choice of routing policy for routing data in the network in a semi-random manner, so that similarly situated packets are not always routed along the same path. Semi-random variation of the routing policy tends to avoid certain local hot spots of network activity, which might otherwise arise using more consistent routing determinations. Preferably, the originating node chooses a routing policy for a packet, and all intermediate nodes in the path route the packet according to that policy. Policies may be rotated on a round-robin basis, selected by generating a random number, or otherwise varied.

Archer, Charles Jens↗

Prediction on Underestimation of Statistical Uncertainty in Monte Carlo Eigenvalue Calculation for Two-dimensional Heterogeneous Color Set Geometry

In a Monte Carlo calculation, statistical uncertainty is inevitably included in the result. The statistical uncertainty is commonly estimated by the central limit theorem without considering correlation among generations. The statistical uncertainty is usually expected to decrease at the inverse square root of the number of neutron histories. However, in a Monte Carlo eigenvalue calculation in a large geometry whose dominance ratio is close to unity, the central limit theorem would not be valid and thus the statistical uncertainty of fission source distribution can be significantly underestimated when inter-cycle correlation is not appropriately taken into account. The true statistical uncertainty can be calculated by statistical processing of the results obtained by multiple Monte Carlo runs using different initial random seeds, but this approach has a drawback of large calculation cost. Therefore, prediction of true statistical uncertainty by alternate approach is desirable. One of the solutions for this issue is an evaluation method of statistical uncertainty based on the eigenfunction expansion and the Autoregressive model (AR model). In this method, the true statistical uncertainty is theoretically derived by applying the AR model to behavior of expansion coefficients of eigenfunction in fission source distribution during cycles. Underestimation ratio of statistical uncertainty of fission source distribution can be predicted by this method. Note that the underestimation ratio of statistical uncertainty is defined by the ratio of estimated error by a Monte Carlo calculation to the true statistical error. In the previous study, the proposed method was verified in a simple condition such as one energy group, one dimensional homogeneous slab geometry. Contributions of this summary are as follows. Firstly, expression of the noise term that is appeared in the AR model is theoretically derived. Secondly, prediction on underestimation ratio of statistical uncertainty is carried out in a more general case, i.e., flux of each fuel cell in 2 x 2 heterogeneous assembly geometry taken from the C5G7 benchmark problem. Through the present study, applicability of the theoretical prediction for underestimation of statistical uncertainty in practical problems will be verified. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

An adaptive Hessian approximated stochastic gradient MCMC method

Bayesian approaches have been successfully integrated into training deep neural networks. One popular family is stochastic gradient Markov chain Monte Carlo methods (SG-MCMC), which have gained increasing interest due to their ability to handle large datasets and the potential to avoid overfitting. Although standard SG-MCMC methods have shown great performance in a variety of problems, they may be inefficient when the random variables in the target posterior densities have scale differences or are highly correlated. Here, we present an adaptive Hessian approximated stochastic gradient MCMC method to incorporate local geometric information while sampling from the posterior. The idea is to apply stochastic approximation (SA) to sequentially update a preconditioning matrix at each iteration. The preconditioner possesses second-order information and can guide the random walk of a sampler efficiently. Instead of computing and saving the full Hessian of the log posterior, we use limited memory of the samples and their stochastic gradients to approximate the inverse Hessian-vector multiplication in the updating formula. Moreover, by smoothly optimizing the preconditioning matrix via SA, our proposed algorithm can asymptotically converge to the target distribution with a controllable bias under mild conditions. To reduce the training and testing computational burden, we adopt a magnitude-based weight pruning method to enforce the sparsity of the network. Our method is user-friendly and demonstrates better learning results compared to standard SG-MCMC updating rules. The approximation of inverse Hessian alleviates storage and computational complexities for large dimensional models. Numerical experiments are performed on several problems, including sampling from 2D correlated distribution, synthetic regression problems, and learning the numerical solutions of heterogeneous elliptic PDE. The numerical results demonstrate great improvement in both the convergence rate and accuracy.

97 MATHEMATICS AND COMPUTING↗

Adaptive sparse polynomial chaos expansion based on least angle regression

Polynomial chaos (PC) expansions are used in stochastic finite element analysis to represent the random model response by a set of coefficients in a suitable (so-called polynomial chaos) basis. The number of terms to be computed grows dramatically with the size of the input random vector, which makes the computational cost of classical solution schemes (may it be intrusive (i.e. of Galerkin type) or non intrusive) unaffordable when the deterministic finite element model is expensive to evaluate. To address such problems, the paper describes a non intrusive method that builds a sparse PC expansion. First, an original strategy for truncating the PC expansions, based on hyperbolic index sets, is proposed. Then an adaptive algorithm based on least angle regression (LAR) is devised for automatically detecting the significant coefficients of the PC expansion. Beside the sparsity of the basis, the experimental design used at each step of the algorithm is systematically complemented in order to avoid the overfitting phenomenon. The accuracy of the PC metamodel is checked using an estimate inspired by statistical learning theory, namely the corrected leave-one-out error. As a consequence, a rather small number of PC terms are eventually retained (sparse representation), which may be obtained at a reduced computational cost compared to the classical 'full' PC approximation. The convergence of the algorithm is shown on an analytical function. Then the method is illustrated on three stochastic finite element problems. The first model features 10 input random variables, whereas the two others involve an input random field, which is discretized into 38 and 30 - 500 random variables, respectively.

97 MATHEMATICS AND COMPUTING↗

Estimating basis functions in massive fields under the spatial mixed effects model

Abstract Spatial prediction is commonly achieved under the assumption of a Gaussian random field by obtaining maximum likelihood estimates of parameters, and then using the kriging equations to arrive at predicted values. For massive datasets, fixed rank kriging using the expectation–maximization algorithm for estimation has been proposed as an alternative to the usual but computationally prohibitive kriging method. The method reduces computation cost of estimation by redefining the spatial process as a linear combination of basis functions and spatial random effects. A disadvantage of this method is that it imposes constraints on the relationship between the observed locations and the knots. We develop an alternative method that utilizes the spatial mixed effects model, but allows for additional flexibility by estimating the range of the spatial dependence between the observations and the knots via an alternating expectation conditional maximization algorithm. Experiments show that our methodology improves estimation without sacrificing prediction accuracy while also minimizing the additional computational burden of extra parameter estimation. The methodology is applied to a temperature dataset archived by the United States National Climate Data Center, with improved results over previous methodology.

Pazdernik, Karl↗

Coherent random lasing in diffusive resonant media

We investigate diffusive propagation of light and consequent random lasing in a medium comprising resonant spherical scatterers. A Monte-Carlo calculation based on photon propagation via three-dimensional random walks is employed to obtain the dwell-times of light in the system. We compare the inter-scatterer and intra-scatterer dwell-times for representative resonant and non-resonant wavelengths. Our results show that more efficient random lasing, with intense coherent modes, is obtained when the gain is present inside the scatterers. Further, a larger reduction in frequency fluctuations is achieved by the system with intra-scatterer gain.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

SU-F-T-600: Influence of Acuros XB and AAA Dose Calculation Algorithms On Plan Quality Metrics and Normal Lung Doses in Lung SBRT

Purpose: To study the influence of superposition-beam model (AAA) and determinant-photon transport-solver (Acuros XB) dose calculation algorithms on the treatment plan quality metrics and on normal lung dose in Lung SBRT. Methods: Treatment plans of 10 Lung SBRT patients were randomly selected. Patients were prescribed to a total dose of 50-54Gy in 3–5 fractions (10?5 or 18?3). Doses were optimized accomplished with 6-MV using 2-arcs (VMAT). Doses were calculated using AAA algorithm with heterogeneity correction. For each plan, plan quality metrics in the categories- coverage, homogeneity, conformity and gradient were quantified. Repeat dosimetry for these AAA treatment plans was performed using AXB algorithm with heterogeneity correction for same beam and MU parameters. Plan quality metrics were again evaluated and compared with AAA plan metrics. For normal lung dose, V{sub 20} and V{sub 5} to (Total lung- GTV) were evaluated. Results: The results are summarized in Supplemental Table 1. PTV volume was mean 11.4 (±3.3) cm{sup 3}. Comparing RTOG 0813 protocol criteria for conformality, AXB plans yielded on average, similar PITV ratio (individual PITV ratio differences varied from −9 to +15%), reduced target coverage (−1.6%) and increased R50% (+2.6%). Comparing normal lung doses, the lung V{sub 20} (+3.1%) and V{sub 5} (+1.5%) were slightly higher for AXB plans compared to AAA plans. High-dose spillage ((V105%PD - PTV)/ PTV) was slightly lower for AXB plans but the % low dose spillage (D2cm) was similar between the two calculation algorithms. Conclusion: AAA algorithm overestimates lung target dose. Routinely adapting to AXB for dose calculations in Lung SBRT planning may improve dose calculation accuracy, as AXB based calculations have been shown to be closer to Monte Carlo based dose predictions in accuracy and with relatively faster computational time. For clinical practice, revisiting dose-fractionation in Lung SBRT to correct for dose overestimates attributable to algorithm may very well be warranted.

60 APPLIED LIFE SCIENCES↗

Efficient computation of basic sums for random polydispersed composites

The main goal of the paper was to develop algorithms and methods for computation of basic sums, the mathematical objects of great importance in computational materials science having applications to description of the representative volume element (RVE) and to the effective properties of 2D composites. The previously used algorithm had the exponential complexity. We propose a linearly complex algorithm. All the presented algorithms can be easily implemented in modern scientific computing packages, while maintaining both efficient calculations and a high level of abstraction. The proposed approach is applied to derivation of a polynomial approximation of the effective conductivity formula for 2D random material with non-overlapping circular inclusions with normally distributed radii. The obtained formulas are applied to the optimal packing problem of disks on the plane.

97 MATHEMATICS AND COMPUTING↗