Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “randomized algorithms”

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
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Randomized Algorithms for Linear Solvers

Recently, randomized algorithms in numerical linear algebra, specifically those centered around random sketching, have gained traction in primarily theoretical research due to their potential to significantly reduce problem dimensionality at the cost of an O(1) multiplicative distortion factor. It has been assumed that this sketching can be done efficiently, but thorough investigation into how precisely to do it has been neglected. Moreover, the theory-based community has argued for sketching’s ability to reduce computational cost via complexity analysis, but has not researched how it affects the stability of the algorithms. At Sandia, efficient linear solvers that scale well on modern HPC architectures while maintaining stability are imperative for practical applications. In this LDRD, we developed a random sketching strategy that is substantially faster than existing ones, and demonstrate its superior performance in practice on a NVIDIA H100 GPU. Moreover, we show how this can be used to significantly outperform existing linear least squares solvers while improving the solver’s stability as well. Additionally, we demonstrate how this sketching strategy can be used to make a fast, stable QR factorization that can subsequently be used in s-step and block Krylov solvers. Finally, we incorporate a sketching-based block orthogonalization scheme into s-step GMRES, which is stable and faster than existing approaches on the Perlmutter supercomputer.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Randomized Algorithms for Low-Rank Matrix and Tensor Decompositions

This paper surveys randomized algorithms in numerical linear algebra for low-rank decompositions of matrices and tensors. The survey begins with a review of classical matrix algorithms that can be accelerated by randomized dimensionality reduction, such as the singular value decomposition (SVD) or interpolative (ID) and CUR decompositions. Recent advances in randomized dimensionality reduction are discussed, including new methods of fast matrix sketching and sampling techniques, which are incorporated into classical matrix algorithms for fast low-rank matrix approximations. The extension of randomized matrix algorithms to tensors is then explored for several low-rank tensor decompositions in the CP and Tucker formats, including the higher-order SVD, ID, and CUR decomposition.

Pearce, Katherine J. [The University of Texas at A↗
DOE OSTI2025

Randomized Algorithms for Symmetric Nonnegative Matrix Factorization

Symmetric Nonnegative Matrix Factorization (SymNMF) is a technique in data analysis and machine learning that approximates a matrix with a product of a nonnegative, low-rank matrix and it transpose. To design faster and more scalable algorithms for SymNMF we develop two randomized algorithms for its computation. The first method uses randomized matrix sketching to compute an initial low-rank approximation to the input matrix and proceeds to uses this as a low-rank input to rapidly compute a SymNMF. The second methods uses randomized leverage score sampling to approximately solve constrained least squares problems. Many successful methods for SymNMF rely on (approximately) solving sequences of constrained least squares problems. Here, we prove theoretically that leverage score sampling can approximately solve constrained least squares problems to e-accuracy. Finally we demonstrate both methods work in practice by applying them to graph clustering tasks on large real world data sets. These experiments show that our methods approximately maintain solution quality and achieve significant speed ups for both large dense and large sparse problems.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS2005

Biased Randomized Algorithm for Fast Model-Based Diagnosis

A biased randomized algorithm has been developed to enable the rapid computational solution of a propositional- satisfiability (SAT) problem equivalent to a diagnosis problem. The closest competing methods of automated diagnosis are described in the preceding article "Fast Algorithms for Model-Based Diagnosis" and "Two Methods of Efficient Solution of the Hitting-Set Problem" (NPO-30584), which appears elsewhere in this issue. It is necessary to recapitulate some of the information from the cited articles as a prerequisite to a description of the present method. As used here, "diagnosis" signifies, more precisely, a type of model-based diagnosis in which one explores any logical inconsistencies between the observed and expected behaviors of an engineering system. The function of each component and the interconnections among all the components of the engineering system are represented as a logical system. Hence, the expected behavior of the engineering system is represented as a set of logical consequences. Faulty components lead to inconsistency between the observed and expected behaviors of the system, represented by logical inconsistencies. Diagnosis - the task of finding the faulty components - reduces to finding the components, the abnormalities of which could explain all the logical inconsistencies. One seeks a minimal set of faulty components (denoted a minimal diagnosis), because the trivial solution, in which all components are deemed to be faulty, always explains all inconsistencies. In the methods of the cited articles, the minimal-diagnosis problem is treated as equivalent to a minimal-hitting-set problem, which is translated from a combinatorial to a computational problem by mapping it onto the Boolean-satisfiability and integer-programming problems. The integer-programming approach taken in one of the prior methods is complete (in the sense that it is guaranteed to find a solution if one exists) and slow and yields a lower bound on the size of the minimal diagnosis. In contrast, the present approach is incomplete and fast and yields an upper bound on the size of the minimal diagnosis.

Williams, Colin↗
DOE OSTI2025

Ensemble Simulation Techniques and Fast Randomized Algorithms

The major goals of the project were to develop and analyze new ensemble simulation techniques, including trajectory stratification and preconditioned MCMC techniques, as well as develop fast numerical linear algebra techniques closely related to ensemble simulation ideas. The trajectory stratification techniques involve simulating in parallel short trajectory fragments of a Markov process confined to a specific region of space‐time and then patching together the statistics gathered to assemble estimates of very general dynamical properties. We have also developed this approach for rare event simulation and extended the techniques to applications requiring a more general framework (such as electronic structure calculations). The preconditioned MCMC techniques involve simulating multiple Markov chains in parallel and then using information from the ensemble to speed the mixing of each individual chain. The fast randomized linear algebra methods are motivated by the diffusion Monte Carlo technique, but are applicable to finding the dominant eigenvalue of (almost) general matrices. For most non‐negative matrices, the schemes result in an error (compared to the power method) that is constant in the dimension of the problem. For more general matrices, we see a very clear sublinear cost trend in computational tests.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2025

Randomized algorithms for accelerating linear algebraic computations

The project supported the development of new methodologies for performing matrix computations that form key building blocks in modern scientific computing, such as low rank approximation of matrices, and efficient representations of global operators that arise in simulations of physical phenomena.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1971

Parameter identification using a creeping-random-search algorithm

A creeping-random-search algorithm is applied to different types of problems in the field of parameter identification. The studies are intended to demonstrate that a random-search algorithm can be applied successfully to these various problems, which often cannot be handled by conventional deterministic methods, and, also, to introduce methods that speed convergence to an extremal of the problem under investigation. Six two-parameter identification problems with analytic solutions are solved, and two application problems are discussed in some detail. Results of the study show that a modified version of the basic creeping-random-search algorithm chosen does speed convergence in comparison with the unmodified version. The results also show that the algorithm can successfully solve problems that contain limits on state or control variables, inequality constraints (both independent and dependent, and linear and nonlinear), or stochastic models.

Parrish, R. V.↗
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↗
NASA NTRS1975

A random search algorithm for laboratory computers

The small laboratory computer is ideal for experimental control and data acquisition. Postexperimental data processing is often performed on large computers because of the availability of sophisticated programs, but costs and data compatibility are negative factors. Parameter optimization can be accomplished on the small computer, offering ease of programming, data compatibility, and low cost. A previously proposed random-search algorithm ('random creep') was found to be very slow in convergence. A method is proposed (the 'random leap' algorithm) which starts in a global search mode and automatically adjusts step size to speed convergence. A FORTRAN executive program for the random-leap algorithm is presented which calls a user-supplied function subroutine. An example of a function subroutine is given which calculates maximum-likelihood estimates of receiver operating-characteristic parameters from binary response data. Other applications in parameter estimation, generalized least squares, and matrix inversion are discussed.

Curry, R. E.↗
NASA NTRS1982

A high speed implementation of the random decrement algorithm

The algorithm is useful for measuring net system damping levels in stochastic processes and for the development of equivalent linearized system response models. The algorithm works by summing together all subrecords which occur after predefined threshold level is crossed. The random decrement signature is normally developed by scanning stored data and adding subrecords together. The high speed implementation of the random decrement algorithm exploits the digital character of sampled data and uses fixed record lengths of 2(n) samples to greatly speed up the process. The contributions to the random decrement signature of each data point was calculated only once and in the same sequence as the data were taken. A hardware implementation of the algorithm using random logic is diagrammed and the process is shown to be limited only by the record size and the threshold crossing frequency of the sampled data. With a hardware cycle time of 200 ns and 1024 point signature, a threshold crossing frequency of 5000 Hertz can be processed and a stably averaged signature presented in real time.

Kiraly, L. J.↗
NASA NTRS1983

A high-speed implementation of the random decrement algorithm

The algorithm is useful for measuring net system damping levels in stochastic processes and for the development of equivalent linearized system response models. The algorithm works by summing together all subrecords which occur after predefined threshold level is crossed. The random decrement signature is normally developed by scanning stored data and adding subrecords together. The high speed implementation of the random decrement algorithm exploits the digital character of sampled data and uses fixed record lengths of 2(n) samples to greatly speed up the process. The contributions to the random decrement signature of each data point was calculated only once and in the same sequence as the data were taken. A hardware implementation of the algorithm using random logic is diagrammed and the process is shown to be limited only by the record size and the threshold crossing frequency of the sampled data. With a hardware cycle time of 200 ns and 1024 point signature, a threshold crossing frequency of 5000 Hertz can be processed and a stably averaged signature presented in real time.

Kiraly, L. J.↗
DOE OSTI2026

Reducing Randomized Quantum Algorithm Cost [Slides]

We derive the optimal sampling strategy for minimizing total resource cost in randomized quantum algorithms. Our framework is completely general, allowing for resources as diverse as gate counts circuit depth, runtime, or even dissipated energy.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Solovay-Kitaev Algorithm and Randomized Compilation Data Availability

This zipped folder contains simulation notebooks, simulated data, and experimental data from the QSCOUT trapped-ion device that were used in the publication "Solovay-Kitaev Algorithm and Randomized Compilation" (https://doi.org/10.1103/ll6m-dbl7). The raw data is in the form of measurement outcomes of simple tomographic quantum circuits that were executed on the QSCOUT device and simulated using JAQALPAQ. These data are used to create plots within the jupyter notebooks that were included in the publication.

Quantum benchmarking↗
NASA NTRS1990

Representation of high frequency Space Shuttle data by ARMA algorithms and random response spectra

High frequency Space Shuttle lift-off data are treated by autoregressive (AR) and autoregressive-moving-average (ARMA) digital algorithms. These algorithms provide useful information on the spectral densities of the data. Further, they yield spectral models which lend themselves to incorporation to the concept of the random response spectrum. This concept yields a reasonably smooth power spectrum for the design of structural and mechanical systems when the available data bank is limited. Due to the non-stationarity of the lift-off event, the pertinent data are split into three slices. Each of the slices is associated with a rather distinguishable phase of the lift-off event, where stationarity can be expected. The presented results are rather preliminary in nature; it is aimed to call attention to the availability of the discussed digital algorithms and to the need to augment the Space Shuttle data bank as more flights are completed.

Spanos, P. D.↗
NASA NTRS2004

SPIN or LURCH : a Comparative Assessment of Model Checking and Stochastic Search for Temporal Properties in Procedural Code

The difficulty of how to test large systems, such as the one on board a NASA robotic remote explorer (RRE) vehicle, is fundamentally a search issue: the global state space representing all possible has yet to be solved, even after many decades of work. Randomized algorithms have been known to outperform their deterministic counterparts for search problems representing a wide range of applications. In the case study presented here, the LURCH randomized algorithm proved to be adequate to the task of testing a NASA RRE vehicle. LURCH found all the errors found by an earlier analysis of a more complete method (SPIN). Our empirical results are that LURCH can scale to much larger models than standard model checkers like SMV and SPIN. Further, the LURCH analysis was simpler than the SPIN analysis. The simplicity and scalability of LURCH are two compelling reasons for experimenting further with this tool.

verification↗
DOE OSTI2025

Karhunen–Loève deep learning method for surrogate modeling and approximate Bayesian parameter estimation

We evaluate the performance of the Karhunen-Loève Deep Neural Network (KL-DNN) framework for surrogate modeling and approximate Bayesian parameter estimation in partial differential equation models. In the surrogate model, the Karhunen-Loève (KL) expansions are used for the dimensionality reduction of the number of unknown parameters and variables, and a deep neural network is employed to relate the reduced space of parameters to that of the state variables. The KL-DNN surrogate model is used to formulate a maximum-a-posteriori-like least-squares problem, which is randomized to draw samples of the posterior distribution of the parameters. We test the proposed framework for a hypothetical unconfined aquifer via comparison with the forward MODFLOW and inverse PEST++ iterative ensemble smoother (IES) solutions as well as the state-of-the-art Fourier neural operator (FNO) and deep operator networks (DeepONets) operator learning surrogate models. Our results show that the KL-DNN surrogate model outperforms FNO and DeepONet for forward predictions. For solving inverse problems, the randomized algorithm provides the same or more accurate Bayesian predictions of the parameters than IES as evidenced by the higher log-predictive probability of both the estimated parameter field and the forecast hydraulic head. The posterior mean obtained from the randomized algorithm is closer to the reference parameter field than that obtained with FNO as the maximum a posteriori estimate.

Approximate Bayesian inference↗