SEARCH · Search NASA

Results for “Vectorized algorithm”

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 217 records · Page 12

All sources NASA NTRS DOE OSTI

NASA NTRS2000

Ordering Unstructured Meshes for Sparse Matrix Computations on Leading Parallel Systems

The ability of computers to solve hitherto intractable problems and simulate complex processes using mathematical models makes them an indispensable part of modern science and engineering. Computer simulations of large-scale realistic applications usually require solving a set of non-linear partial differential equations (PDES) over a finite region. For example, one thrust area in the DOE Grand Challenge projects is to design future accelerators such as the SpaHation Neutron Source (SNS). Our colleagues at SLAC need to model complex RFQ cavities with large aspect ratios. Unstructured grids are currently used to resolve the small features in a large computational domain; dynamic mesh adaptation will be added in the future for additional efficiency. The PDEs for electromagnetics are discretized by the FEM method, which leads to a generalized eigenvalue problem Kx = AMx, where K and M are the stiffness and mass matrices, and are very sparse. In a typical cavity model, the number of degrees of freedom is about one million. For such large eigenproblems, direct solution techniques quickly reach the memory limits. Instead, the most widely-used methods are Krylov subspace methods, such as Lanczos or Jacobi-Davidson. In all the Krylov-based algorithms, sparse matrix-vector multiplication (SPMV) must be performed repeatedly. Therefore, the efficiency of SPMV usually determines the eigensolver speed. SPMV is also one of the most heavily used kernels in large-scale numerical simulations.

Oliker, Leonid↗

NASA NTRS2006

Active Learning with Irrelevant Examples

Active learning algorithms attempt to accelerate the learning process by requesting labels for the most informative items first. In real-world problems, however, there may exist unlabeled items that are irrelevant to the user's classification goals. Queries about these points slow down learning because they provide no information about the problem of interest. We have observed that when irrelevant items are present, active learning can perform worse than random selection, requiring more time (queries) to achieve the same level of accuracy. Therefore, we propose a novel approach, Relevance Bias, in which the active learner combines its default selection heuristic with the output of a simultaneously trained relevance classifier to favor items that are likely to be both informative and relevant. In our experiments on a real-world problem and two benchmark datasets, the Relevance Bias approach significantly improved the learning rate of three different active learning approaches.

machine learning↗

NASA NTRS2007

Automated Knowledge Discovery From Simulators

A computational method, SimLearn, has been devised to facilitate efficient knowledge discovery from simulators. Simulators are complex computer programs used in science and engineering to model diverse phenomena such as fluid flow, gravitational interactions, coupled mechanical systems, and nuclear, chemical, and biological processes. SimLearn uses active-learning techniques to efficiently address the "landscape characterization problem." In particular, SimLearn tries to determine which regions in "input space" lead to a given output from the simulator, where "input space" refers to an abstraction of all the variables going into the simulator, e.g., initial conditions, parameters, and interaction equations. Landscape characterization can be viewed as an attempt to invert the forward mapping of the simulator and recover the inputs that produce a particular output. Given that a single simulation run can take days or weeks to complete even on a large computing cluster, SimLearn attempts to reduce costs by reducing the number of simulations needed to effect discoveries. Unlike conventional data-mining methods that are applied to static predefined datasets, SimLearn involves an iterative process in which a most informative dataset is constructed dynamically by using the simulator as an oracle. On each iteration, the algorithm models the knowledge it has gained through previous simulation trials and then chooses which simulation trials to run next. Running these trials through the simulator produces new data in the form of input-output pairs. The overall process is embodied in an algorithm that combines support vector machines (SVMs) with active learning. SVMs use learning from examples (the examples are the input-output pairs generated by running the simulator) and a principle called maximum margin to derive predictors that generalize well to new inputs. In SimLearn, the SVM plays the role of modeling the knowledge that has been gained through previous simulation trials. Active learning is used to determine which new input points would be most informative if their output were known. The selected input points are run through the simulator to generate new information that can be used to refine the SVM. The process is then repeated. SimLearn carefully balances exploration (semi-randomly searching around the input space) versus exploitation (using the current state of knowledge to conduct a tightly focused search). During each iteration, SimLearn uses not one, but an ensemble of SVMs. Each SVM in the ensemble is characterized by different hyper-parameters that control various aspects of the learned predictor - for example, whether the predictor is constrained to be very smooth (nearby points in input space lead to similar output predictions) or whether the predictor is allowed to be "bumpy." The various SVMs will have different preferences about which input points they would like to run through the simulator next. SimLearn includes a formal mechanism for balancing the ensemble SVM preferences so that a single choice can be made for the next set of trials.

Burl, Michael↗

NASA NTRS2010

Multi-Stage System for Automatic Target Recognition

A multi-stage automated target recognition (ATR) system has been designed to perform computer vision tasks with adequate proficiency in mimicking human vision. The system is able to detect, identify, and track targets of interest. Potential regions of interest (ROIs) are first identified by the detection stage using an Optimum Trade-off Maximum Average Correlation Height (OT-MACH) filter combined with a wavelet transform. False positives are then eliminated by the verification stage using feature extraction methods in conjunction with neural networks. Feature extraction transforms the ROIs using filtering and binning algorithms to create feature vectors. A feedforward back-propagation neural network (NN) is then trained to classify each feature vector and to remove false positives. The system parameter optimizations process has been developed to adapt to various targets and datasets. The objective was to design an efficient computer vision system that can learn to detect multiple targets in large images with unknown backgrounds. Because the target size is small relative to the image size in this problem, there are many regions of the image that could potentially contain the target. A cursory analysis of every region can be computationally efficient, but may yield too many false positives. On the other hand, a detailed analysis of every region can yield better results, but may be computationally inefficient. The multi-stage ATR system was designed to achieve an optimal balance between accuracy and computational efficiency by incorporating both models. The detection stage first identifies potential ROIs where the target may be present by performing a fast Fourier domain OT-MACH filter-based correlation. Because threshold for this stage is chosen with the goal of detecting all true positives, a number of false positives are also detected as ROIs. The verification stage then transforms the regions of interest into feature space, and eliminates false positives using an artificial neural network classifier. The multi-stage system allows tuning the detection sensitivity and the identification specificity individually in each stage. It is easier to achieve optimized ATR operation based on its specific goal. The test results show that the system was successful in substantially reducing the false positive rate when tested on a sonar and video image datasets.

Chao, Tien-Hsin↗

NASA NTRS2012

Improved Hierarchical Optimization-Based Classification of Hyperspectral Images Using Shape Analysis

A new spectral-spatial method for classification of hyperspectral images is proposed. The HSegClas method is based on the integration of probabilistic classification and shape analysis within the hierarchical step-wise optimization algorithm. First, probabilistic support vector machines classification is applied. Then, at each iteration two neighboring regions with the smallest Dissimilarity Criterion (DC) are merged, and classification probabilities are recomputed. The important contribution of this work consists in estimating a DC between regions as a function of statistical, classification and geometrical (area and rectangularity) features. Experimental results are presented on a 102-band ROSIS image of the Center of Pavia, Italy. The developed approach yields more accurate classification results when compared to previously proposed methods.

Classification↗

NASA NTRS2009

Neural Network Target Identification System for False Alarm Reduction

ATR↗

NASA NTRS2019

IMERG V06: Changes to the Morphing Algorithm

As the US Science Team’s globally gridded precipitation product from the NASA/JAXA Global Precipitation Measurement (GPM) mission, the Integrated Multi-satellitE Retrievals for GPM (IMERG) estimates the surface precipitation rates at 0.1° every half-hour using spaceborne sensors for various scientific and societal applications. One key component of IMERG is the morphing algorithm, which uses motion vectors to perform quasi-Lagrangian interpolation to fill in gaps in the passive microwave precipitation field using motion vectors. Up to IMERG V05, the motion vectors were derived from the large-scale motions of infrared observations of cloud tops. This study details the changes introduced in IMERG V06 to derive motion vectors from large-scale motions of selected atmospheric variables in numerical models, which allow IMERG estimates to be extended from the 60°N/S latitude band to the entire globe. Evaluation against both instantaneous passive microwave retrievals and ground measurements demonstrates the general improvement in the precipitation field of the new approach. Most of the model variables tested exhibited similar performance, but total precipitable water vapor was chosen as the source of the motion vectors for IMERG V06 due to its competitive performance and global completeness. Continuing assessments will provide further insights into possible refinements of this revised morphing scheme in future versions of IMERG.

Jackson Tan↗

NASA NTRS1999

Multi-Parent Clustering Algorithms from Stochastic Grammar Data Models

We introduce a statistical data model and an associated optimization-based clustering algorithm which allows data vectors to belong to zero, one or several parent clusters.

Clustering↗

NASA NTRS1991

Path planning for assembly of strut-based structures

A path planning method with collision avoidance for a general single chain nonredundant or redundant robot is proposed. Joint range boundary overruns are also avoided. The result is a sequence of joint vectors which are passed to a trajectory planner. A potential field algorithm in joint space computes incremental joint vectors delta-q = delta-q(sub a) + delta-q(sub c) + delta-q(sub r). Adding delta-q to the robot's current joint vector leads to the next step in the path. Delta-q(sub a) is obtained by computing the minimum norm solution of the underdetermined linear system J delta-q(sub a) = x(sub a) where x(sub a) is a translational and rotational force vector that attracts the robot to its goal position and orientation. J is the manipulator Jacobian. Delta-q(sub c) is a collision avoidance term encompassing collisions between the robot (links and payload) and obstacles in the environment as well as collisions among links and payload of the robot themselves. It is obtained in joint space directly. Delta-q(sub r) is a function of the current joint vector and avoids joint range overruns. A higher level discrete search over candidate safe positions is used to provide alternatives in case the potential field algorithm encounters a local minimum and thus fails to reach the goal. The best first search algorithm A* is used for graph search. Symmetry properties of the payload and equivalent rotations are exploited to further enlarge the number of alternatives passed to the potential field algorithm.

Muenger, Rolf↗

NASA NTRS1991

Vectorization of a Monte Carlo simulation scheme for nonequilibrium gas dynamics

Significant improvement has been obtained in the numerical performance of a Monte Carlo scheme for the analysis of nonequilibrium gas dynamics through an implementation of the algorithm which takes advantage of vector hardware, as presently demonstrated through application to three different problems. These are (1) a 1D standing-shock wave; (2) the flow of an expanding gas through an axisymmetric nozzle; and (3) the hypersonic flow of Ar gas over a 3D wedge. Problem (3) is illustrative of the greatly increased number of molecules which the simulation may involve, thanks to improved algorithm performance.

Boyd, Iain D.↗

NASA NTRS1993

Rate determination from vector observations

Vector observations are a common class of attitude data provided by a wide variety of attitude sensors. Attitude determination from vector observations is a well-understood process and numerous algorithms such as the TRIAD algorithm exist. These algorithms require measurement of the line of site (LOS) vector to reference objects and knowledge of the LOS directions in some predetermined reference frame. Once attitude is determined, it is a simple matter to synthesize vehicle rate using some form of lead-lag filter, and then, use it for vehicle stabilization. Many situations arise, however, in which rate knowledge is required but knowledge of the nominal LOS directions are not available. This paper presents two methods for determining spacecraft angular rates from vector observations without a priori knowledge of the vector directions. The first approach uses an extended Kalman filter with a spacecraft dynamic model and a kinematic model representing the motion of the observed LOS vectors. The second approach uses a 'differential' TRIAD algorithm to compute the incremental direction cosine matrix, from which vehicle rate is then derived.

Weiss, Jerold L.↗

NASA NTRS1991

Krylov model reduction algorithm for undamped structural dynamics systems

Krylov vectors furnish an efficient basis for eigenvalue analysis and model reduction of structural dynamics systems. The reduced-order model obtained by the present Krylov model-reduction algorithm for an undamped structural-dynamics system is found to match low-frequency moments. The transformed system equation in Krylov coordinates reflects the structure of a tandem system.

Craig, Roy R., Jr.↗

NASA NTRS1995

A comparison of model-based VQ compression with other VQ approaches

In our previous work on Model-Based Vector Quantization (MVQ), we presented some performance comparisons (both rate distortion and decompression time) with VQ and JPEG/DCT. In this paper, we compare the MVQ's rate distortion performance with Mean Removed Vector Quantization (MRVQ) and include our previous comparison with VQ. MVQ is similar to MRVQ in many ways. Both of these techniques extract means of the vectors (raster-scanned image blocks) and reduce them to mean removed residuals by subtracting block means from the elements of the vectors. In the case of MRVQ, a codebook of residual vectors is generated using a training set. For every vector from the input image, the block mean and address of the codevector from the codebook that matches the input vector closest are transmitted to the decoder. The codebook is generated using generalized Lloyd algorithm on training set of residual vectors. For MVQ the pairs consist of vector means and address of the closest matching vector from codebook generated by models based on statistical properties of the residuals and Human Visual System (HVS). In our experiments, we found that MVQ performance in rate distortion sense is almost always better than VQ and is comparable to MRVQ. Further, MVQ is much easier to use than either VQ or MRVQ, since the training and managing of explicit codebooks is not required.

Manohar, Mareboyana↗

NASA NTRS2002

Evaluation of a Pair-Wise Conflict Detection and Resolution Algorithm in a Multiple Aircraft Scenario

The KB3D algorithm is a pairwise conflict detection and resolution (CD&R) algorithm. It detects and generates trajectory vectoring for an aircraft which has been predicted to be in an airspace minima violation within a given look-ahead time. It has been proven, using mechanized theorem proving techniques, that for a pair of aircraft, KB3D produces at least one vectoring solution and that all solutions produced are correct. Although solutions produced by the algorithm are mathematically correct, they might not be physically executable by an aircraft or might not solve multiple aircraft conflicts. This paper describes a simple solution selection method which assesses all solutions generated by KB3D and determines the solution to be executed. The solution selection method and KB3D are evaluated using a simulation in which N aircraft fly in a free-flight environment and each aircraft in the simulation uses KB3D to maintain separation. Specifically, the solution selection method filters KB3D solutions which are procedurally undesirable or physically not executable and uses a predetermined criteria for selection.

Carreno, Victor A.↗

NASA NTRS1988

Robust vector quantization for noisy channels

The paper briefly discusses techniques for making vector quantizers more tolerant to tranmsission errors. Two algorithms are presented for obtaining an efficient binary word assignment to the vector quantizer codewords without increasing the transmission rate. It is shown that about 4.5 dB gain over random assignment can be achieved with these algorithms. It is also proposed to reduce the effects of error propagation in vector-predictive quantizers by appropriately constraining the response of the predictive loop. The constrained system is shown to have about 4 dB of SNR gain over an unconstrained system in a noisy channel, with a small loss of clean-channel performance.

Demarca, J. R. B.↗

NASA NTRS1993

Implementation of the block-Krylov boundary flexibility method of component synthesis

A method of dynamic substructuring is presented which utilizes a set of static Ritz vectors as a replacement for normal eigenvectors in component mode synthesis. This set of Ritz vectors is generated in a recurrence relationship, which has the form of a block-Krylov subspace. The initial seed to the recurrence algorithm is based on the boundary flexibility vectors of the component. This algorithm is not load-dependent, is applicable to both fixed and free-interface boundary components, and results in a general component model appropriate for any type of dynamic analysis. This methodology was implemented in the MSC/NASTRAN normal modes solution sequence using DMAP. The accuracy is found to be comparable to that of component synthesis based upon normal modes. The block-Krylov recurrence algorithm is a series of static solutions and so requires significantly less computation than solving the normal eigenspace problem.

Carney, Kelly S.↗

NASA NTRS1993

Polar decomposition for attitude determination from vector observations

This work treats the problem of weighted least squares fitting of a 3D Euclidean-coordinate transformation matrix to a set of unit vectors measured in the reference and transformed coordinates. A closed-form analytic solution to the problem is re-derived. The fact that the solution is the closest orthogonal matrix to some matrix defined on the measured vectors and their weights is clearly demonstrated. Several known algorithms for computing the analytic closed form solution are considered. An algorithm is discussed which is based on the polar decomposition of matrices into the closest unitary matrix to the decomposed matrix and a Hermitian matrix. A somewhat longer improved algorithm is suggested too. A comparison of several algorithms is carried out using simulated data as well as real data from the Upper Atmosphere Research Satellite. The comparison is based on accuracy and time consumption. It is concluded that the algorithms based on polar decomposition yield a simple although somewhat less accurate solution. The precision of the latter algorithms increase with the number of the measured vectors and with the accuracy of their measurement.

Bar-Itzhack, Itzhack Y.↗

NASA NTRS1996

Optimized TRIAD Algorithm for Attitude Determination

TRIAD is a well known simple algorithm that generates the attitude matrix between two coordinate systems when the components of two abstract vectors are given in the two systems. TRIAD however, is sensitive to the order in which the algorithm handles the vectors, such that the resulting attitude matrix is influenced more by the vector processed first. In this work we present a new algorithm, which we call Optimized TRIAD, that blends in a specified manner the two matrices generated by TRIAD when processing one vector first, and then when processing the other vector first. On the average, Optimized TRIAD yields a matrix which is better than either one of the two matrices in that is ti the closest to the correct matrix. This result is demonstrated through simulation.

Bar-Itzhack, Itzhack Y.↗