SEARCH · Search NASA

Results for “generalized least squared”

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.

SourceResearch connectionFrom yearThrough year

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

At least 109 records · Page 6

All sources NASA NTRS DOE OSTI

DOE OSTI2025

Surrogate-Based Autotuning for Randomized Sketching Algorithms in Regression Problems

Algorithms from Randomized Numerical Linear Algebra (RandNLA) are known to be effective in handling high-dimensional computational problems, providing high-quality empirical performance as well as strong probabilistic guarantees. However, their practical application is complicated by the fact that the user needs to set various algorithm-specific tuning parameters which are different from those used in traditional NLA. This paper demonstrates how a surrogate-based autotuning approach can be used to address fundamental problems of parameter selection in RandNLA algorithms. In particular, we provide a detailed investigation of surrogate-based autotuning for sketch-and-precondition (SAP)-based randomized least squares methods, which have been one of the great success stories in modern RandNLA. Empirical results show that our surrogate-based autotuning approach can achieve near-optimal performance with much less tuning cost than a random search (up to about 7.6x fewer trials of different parameter configurations). Moreover, while our experiments focus on least squares, our results demonstrate a general-purpose autotuning pipeline applicable to any kind of RandNLA algorithm.

Cho, Younghyun↗

DOE OSTI2022

Deterministic and Monte Carlo Nuclear Data Adjustment Methods [Slides]

For the Bayesian Monte Carlo methodology, a need to understand convergence of the posterior moments as a function of the number of parameter realizations is required. In high-dimensional systems, it can be very costly to sample entire parameter space and perform functional evaluation for every realization. Bayesian Monte Carlo allows one to relax the GLLS approximations of model linearity and prior/posterior PDF shape. The Bayesian Stochastic Collocation Method is a deterministic approach to “sample” the parameter space. It allows one to relax the GLLS approximations of model linearity and posterior PDF shape. Higher-order posterior moments (i.e., skewness, kurtosis, etc.) can be studied through polynomial expansion. Tensor product quadrature scales poorly and can use sparse grid quadrature methods.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

DOE OSTI2021

ACCRUE—An Integral Index for Measuring Experimental Relevance in Support of Neutronic Model Validation

A key challenge for the introduction of any design changes, e.g., advanced fuel concepts, first-of-a-kind nuclear reactor designs, etc., is the cost of the associated experiments, which are required by law to validate the use of computer models for the various stages, starting from conceptual design, to deployment, licensing, operation, and safety. To achieve that, a criterion is needed to decide on whether a given experiment, past or planned, is relevant to the application of interest. This allows the analyst to select the best experiments for the given application leading to the highest measures of confidence for the computer model predictions. The state-of-the-art methods rely on the concept of similarity or representativity, which is a linear Gaussian-based inner-product metric measuring the angle—as weighted by a prior model parameters covariance matrix—between two gradients, one representing the application and the other a single validation experiment. This manuscript emphasizes the concept of experimental relevance which extends the basic similarity index to account for the value accrued from past experiments and the associated experimental uncertainties, both currently missing from the extant similarity methods. Accounting for multiple experiments is key to the overall experimental cost reduction by prescreening for redundant information from multiple equally-relevant experiments as measured by the basic similarity index. Accounting for experimental uncertainties is also important as it allows one to select between two different experimental setups, thus providing for a quantitative basis for sensor selection and optimization. The proposed metric is denoted by ACCRUE, short for Accumulative Correlation Coefficient for Relevance of Uncertainties in Experimental validation. Using a number of criticality experiments for highly enriched fast metal systems and low enriched thermal compound systems with accident tolerant fuel concept, the manuscript will compare the performance of the ACCRUE and basic similarity indices for prioritizing the relevance of a group of experiments to the given application.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS1972

Flexible rotor balancing by the influence coefficient method. Part 1: Evaluation of the exact point-speed and least squares procedures

The practical aspects of balancing real, flexible rotors were investigated through inclusion of rotor out-of-roundness data at the measurement probe locations. The computer program was demonstrated to be fully capable of handling out-of-roundness data in the investigation. Testing was performed predominantly with a machine having a 41-inch long, 126-pound rotor. This rotor was operated over a speed range encompassing three rotor-bearing system critical speeds. Both balancing procedures were evaluated for several different conditions of initial rotor unbalance. Safe (and slow) passage through all the critical speeds was obtained after two or three balancing runs in most cases. The Least Squares procedure was found to be generally equivalent in capability to the Exact Point-Speed procedure for the configurations studies. (U)

Tessarzik, J. M.↗

NASA NTRS2021

PolyFit: A C++ code for Polynomial Curve Fit with Calculation of Error Bars

In radiobiology, many dose-response results are modeled using the so-called linear-quadratic (LQ)model, which means that results are modeled as a function of dose Das 𝑅(𝐷)=𝛽0+𝛽1𝐷+𝛽2𝐷2. The coefficients 𝛽0, 𝛽1and 𝛽2are obtained from fitting a series of data points (𝑥𝑖, 𝑦𝑖), which is usually done using a least-square method. The LQ and more generally the polynomial fit capability is implemented in many software that analyzes data. However, there are some instances where the fitting needs to be done programmatically. Furthermore, depending on the software used, some features may not be implemented. In this mini-review, I discuss the basis of polynomial fitting, including the calculation of errors on the coefficients and results, use of weighting and fixing the intercept value (the coefficient 𝛽0). A simple C++ code to perform the polynomial curve fitting is also provided. This code should be useful not only in radiobiology but in other fields of science as well.

Ianik Plante↗

NASA NTRS2021

PolyFit: A C++ code for Polynomial Curve Fit with Calculation of Error Bars

In radiobiology, many dose-response results are modeled using the so-called linear-quadratic(LQ)model, which means that results are modeled as a function of dose Das 𝑅(𝐷)=𝛽0+𝛽1𝐷+𝛽2𝐷2. The coefficients 𝛽0, 𝛽1and 𝛽2are obtained from fitting a series of data points (𝑥𝑖,𝑦𝑖), which is usually done using a least-square method. The LQ and more generally the polynomial fit capability is implemented in many software that analyzes data. However, it is often convenient to do the fitting programmatically, especially when a large number of datasets should be analyzed. Furthermore, depending on the software used, some features may not be implemented. In this mini-review, I discuss the basis of polynomial fitting, including the calculation of errors on the coefficients and results, use of weighting and fixing the intercept value (the coefficient 𝛽0). A simple C++ code to perform the polynomial curve fitting is also provided. This code should be useful not only in radiobiology but in other fields of science as well.

Ianik Plante↗

NASA NTRS1994

A simple suboptimal least-squares algorithm for attitude determination with multiple sensors

Three-axis attitude determination is equivalent to finding a coordinate transformation matrix which transforms a set of reference vectors fixed in inertial space to a set of measurement vectors fixed in the spacecraft. The attitude determination problem can be expressed as a constrained optimization problem. The constraint is that a coordinate transformation matrix must be proper, real, and orthogonal. A transformation matrix can be thought of as optimal in the least-squares sense if it maps the measurement vectors to the reference vectors with minimal 2-norm errors and meets the above constraint. This constrained optimization problem is known as Wahba's problem. Several algorithms which solve Wahba's problem exactly have been developed and used. These algorithms, while steadily improving, are all rather complicated. Furthermore, they involve such numerically unstable or sensitive operations as matrix determinant, matrix adjoint, and Newton-Raphson iterations. This paper describes an algorithm which minimizes Wahba's loss function, but without the constraint. When the constraint is ignored, the problem can be solved by a straightforward, numerically stable least-squares algorithm such as QR decomposition. Even though the algorithm does not explicitly take the constraint into account, it still yields a nearly orthogonal matrix for most practical cases; orthogonality only becomes corrupted when the sensor measurements are very noisy, on the same order of magnitude as the attitude rotations. The algorithm can be simplified if the attitude rotations are small enough so that the approximation sin(theta) approximately equals theta holds. We then compare the computational requirements for several well-known algorithms. For the general large-angle case, the QR least-squares algorithm is competitive with all other know algorithms and faster than most. If attitude rotations are small, the least-squares algorithm can be modified to run faster, and this modified algorithm is faster than all but a similarly specialized version of the QUEST algorithm. We also introduce a novel measurement averaging technique which reduces the n-measurement case to the two measurement case for our particular application, a star tracker and earth sensor mounted on an earth-pointed geosynchronous communications satellite. Using this technique, many n-measurement problems reduce to less than or equal to 3 measurements; this reduces the amount of required calculation without significant degradation in accuracy. Finally, we present the results of some tests which compare the least-squares algorithm with the QUEST and FOAM algorithms in the two-measurement case. For our example case, all three algorithms performed with similar accuracy.

Brozenec, Thomas F.↗

DOE OSTI2021

Efficient ptychographic phase retrieval via a matrix-free Levenberg-Marquardt algorithm

The phase retrieval problem, where one aims to recover a complex-valued image from far-field intensity measurements, is a classic problem encountered in a range of imaging applications. Modern phase retrieval approaches usually rely on gradient descent methods in a nonlinear minimization framework. Calculating closed-form gradients for use in these methods is tedious work, and formulating second order derivatives is even more laborious. Additionally, second order techniques often require the storage and inversion of large matrices of partial derivatives, with memory requirements that can be prohibitive for data-rich imaging modalities. We use a reverse-mode automatic differentiation (AD) framework to implement an efficient matrix-free version of the Levenberg-Marquardt (LM) algorithm, a longstanding method that finds popular use in nonlinear least-square minimization problems but which has seen little use in phase retrieval. Furthermore, we extend the basic LM algorithm so that it can be applied for more general constrained optimization problems (including phase retrieval problems) beyond just the least-square applications. Since we use AD, we only need to specify the physics-based forward model for a specific imaging application; the first and second-order derivative terms are calculated automatically through matrix-vector products, without explicitly forming the large Jacobian or Gauss-Newton matrices typically required for the LM method. We demonstrate that this algorithm can be used to solve both the unconstrained ptychographic object retrieval problem and the constrained “blind” ptychographic object and probe retrieval problems, under the popular Gaussian noise model as well as the Poisson noise model. We compare this algorithm to state-of-the-art first order ptychographic reconstruction methods to demonstrate empirically that this method outperforms best-in-class first-order methods: it provides excellent convergence guarantees with (in many cases) a superlinear rate of convergence, all with a computational cost comparable to, or lower than, the tested first-order algorithms.

36 MATERIALS SCIENCE↗

NASA NTRS1981

On the method of least squares. II

A least squares method proposed previously for solving a general class of problems is expanded in two ways. First, covariance matrices related to the solution are calculated and their interpretation is given. Second, improved methods of solving the normal equations related to those of Marquardt (1963) and Fletcher and Powell (1963) are developed for this approach. These methods may converge in cases where Newton's method diverges or converges slowly.

Jefferys, W. H.↗

NASA NTRS1991

Computer Processing Of Tunable-Diode-Laser Spectra

Tunable-diode-laser spectrometer measuring transmission spectrum of gas operates under control of computer, which also processes measurement data. Measurements in three channels processed into spectra. Computer controls current supplied to tunable diode laser, stepping it through small increments of wavelength while processing spectral measurements at each step. Program includes library of routines for general manipulation and plotting of spectra, least-squares fitting of direct-transmission and harmonic-absorption spectra, and deconvolution for determination of laser linewidth and for removal of instrumental broadening of spectral lines.

May, Randy D.↗

NASA NTRS1990

Improved LMS algorithm for adaptive beamforming

Two adaptive algorithms which make use of all the available samples to estimate the required gradient are proposed and studied. The first algorithm is referred to as the recursive LMS (least mean squares) and is applicable to a general array. The second algorithm is referred to as the improved LMS algorithm and exploits the Toeplitz structure of the ACM (array correlation matrix); it can be used only for an equispaced linear array.

Godara, Lal C.↗

NASA NTRS2005

New Algorithms for Estimating Spacecraft Position Using Scanning Techniques for Deep Space Network Antennas

As more and more nonlinear estimation techniques become available, our interest is in finding out what performance improvement, if any, they can provide for practical nonlinear problems that have been traditionally solved using linear methods. In this paper we examine the problem of estimating spacecraft position using conical scan (conscan) for NASA's Deep Space Network antennas. We show that for additive disturbances on antenna power measurement, the problem can be transformed into a linear one, and we present a general solution to this problem, with the least square solution reported in literature as a special case. We also show that for additive disturbances on antenna position, the problem is a truly nonlinear one, and we present two approximate solutions based on linearization and Unscented Transformation respectively, and one 'exact' solution based on Markov Chain Monte Carlo (MCMC) method. Simulations show that, with the amount of data collected in practice, linear methods perform almost the same as MCMC methods. It is only when we artificially reduce the amount of collected data and increase the level of noise that nonlinear methods show significantly better accuracy than that achieved by linear methods, at the expense of more computation.

Metropolis-Hastings (MH)↗

NASA NTRS2014

Generalized Linear Covariance Analysis

This talk presents a comprehensive approach to filter modeling for generalized covariance analysis of both batch least-squares and sequential estimators. We review and extend in two directions the results of prior work that allowed for partitioning of the state space into solve-for'' and consider'' parameters, accounted for differences between the formal values and the true values of the measurement noise, process noise, and textita priori solve-for and consider covariances, and explicitly partitioned the errors into subspaces containing only the influence of the measurement noise, process noise, and solve-for and consider covariances. In this work, we explicitly add sensitivity analysis to this prior work, and relax an implicit assumption that the batch estimator's epoch time occurs prior to the definitive span. We also apply the method to an integrated orbit and attitude problem, in which gyro and accelerometer errors, though not estimated, influence the orbit determination performance. We illustrate our results using two graphical presentations, which we call the variance sandpile'' and the sensitivity mosaic,'' and we compare the linear covariance results to confidence intervals associated with ensemble statistics from a Monte Carlo analysis.

Navagation↗

NASA NTRS2014

Generalized Linear Covariance Analysis

n/a↗

NASA NTRS1983

A new algorithm for constrained nonlinear least-squares problems, part 1

A Gauss-Newton algorithm is presented for solving nonlinear least squares problems. The problem statement may include simple bounds or more general constraints on the unknowns. The algorithm uses a trust region that allows the objective function to increase with logic for retreating to best values. The computations for the linear problem are done using a least squares system solver that allows for simple bounds and linear constraints. The trust region limits are defined by a box around the current point. In its current form the algorithm is effective only for problems with small residuals, linear constraints and dense Jacobian matrices. Results on a set of test problems are encouraging.

Hanson, R. J.↗

NASA NTRS1976

On the interpretation of least squares collocation

A demonstration is given of the strict mathematical equivalence between the least squares collocation and the classical minimum variance estimates. It is shown that the least squares collocation algorithms are a special case of the modified minimum variance estimates. The computational efficiency of several forms of the general minimum variance estimation algorithm is discussed. It is pointed out that for certain geodetic applications the least square collocation algorithm may provide a more efficient formulation of the results from the point of view of the computations required.

Tapley, B. D.↗

DOE OSTI2025

Extended Galerkin Neural Network Approximation of Singular Variational Problems with Error Control

We present extended Galerkin neural networks, a variational framework for approximating general boundary value problems (BVPs) with error control. The main contributions of this work are (1) a rigorous theory guiding the construction of new weighted least squares variational formulations suitable for use in neural network approximation of general BVPs, and (2) an “extended” feedforward network architecture which incorporates and is even capable of learning singular solution structures, thus greatly improving approximability of singular solutions. Furthermore, numerical results are presented for several problems, including steady Stokes flow around reentrant corners and in convex corners with Moffatt eddies in order to demonstrate efficacy of the method.

a posteriori error estimate↗

NASA NTRS1972

Neutron die-away experiment for remote analysis of the surface of the moon and the planets, phase 3

Continuing work on the two die-away measurements proposed to be made in the combined pulsed neutron experiment (CPNE) for analysis of lunar and planetary surfaces is described. This report documents research done during Phase 3. A general exposition of data analysis by the least-squares method and the related problem of the prediction of variance is given. A data analysis procedure for epithermal die-away data has been formulated. In order to facilitate the analysis, the number of independent material variables has been reduced to two: the hydrogen density and an effective oxygen density, the latter being determined uniquely from the nonhydrogeneous elemental composition. Justification for this reduction in the number of variables is based on a set of 27 new theoretical calculations. Work is described related to experimental calibration of the epithermal die-away measurement. An interim data analysis technique based solely on theoretical calculations seems to be adequate and will be used for future CPNE field tests.

Mills, W. R.↗