A Comparison of Several Bandwidth and Profile Reduction Algorithms

Gibbs, Norman E.; Poole, William G.; Stockmeyer, Paul K.

doi:10.1145/355705.355707

Cited by 87 publications

(28 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Equation (4) is a second-order ordinary differential equation in time and is discretized using the second-order Newmark scheme (Belytschko et al, 2000) wherein a banded LU decomposition is used to solve the system of algebraic equations. The Cuthill-McKee (Cuthill and Mckee, 1969) and Gibbs-Poole-Stockmeyer methods (Gibbs et al, 1976) are used to re-index the nodes in order to create banded matrixes. The details of this solid solver can be found in Zheng (2009).…”

Section: B Vocal Fold Tissue Modelingmentioning

confidence: 99%

Direct-numerical simulation of the glottal jet and vocal-fold dynamics in a three-dimensional laryngeal model

Zheng

Mittal

Xue

et al. 2011

The Journal of the Acoustical Society of America

View full text Add to dashboard Cite

An immersed-boundary method based flow solver coupled with a finite-element solid dynamics solver is employed in order to conduct direct-numerical simulations of phonatory dynamics in a three-dimensional model of the human larynx. The computed features of the glottal flow including mean and peak flow rates, and the open and skewness quotients are found to be within the normal physiological range. The flow-induced vibration pattern shows the classical "convergent-divergent" glottal shape, and the vibration amplitude is also found to be typical for human phonation. The vocal fold motion is analyzed through the method of empirical eigenfunctions and this analysis indicates a 1:1 modal entrainment between the "adduction-abduction" mode and the "mucosal wave" mode. The glottal jet is found to exhibit noticeable cycle-to-cycle asymmetric deflections and the mechanism underlying this phenomenon is examined.

show abstract

Section: B Vocal Fold Tissue Modelingmentioning

confidence: 99%

Direct-numerical simulation of the glottal jet and vocal-fold dynamics in a three-dimensional laryngeal model

Zheng

Mittal

Xue

et al. 2011

The Journal of the Acoustical Society of America

View full text Add to dashboard Cite

show abstract

“…Thus, our aim is to find a second permutation P 2 such that S 2 = P 2 S 1 P ⊤ 2 has as much mass as possible in its block-super and sub diagonals, i.e., as many of the inter cluster interactions as possible are accounted for. Finding such permutations is studied in the literature on Bandwidth Minimization and is known to be NP-Hard [9]. We propose a greedy heuristic suitable for bundle adjustment.…”

Section: Cluster-tridiagonalmentioning

confidence: 99%

Visibility Based Preconditioning for bundle adjustment

Kushal

Agarwal

2012

2012 IEEE Conference on Computer Vision and Pattern Recognition

View full text Add to dashboard Cite

We present Visibility Based Preconditioning (VBP) a new technique for efficiently solving the linear least squares problems that arise in bundle adjustment [24]. Using the camera-point visibility structure of the scene, we describe the construction of two preconditioners. These preconditioners when combined with an inexact step LevenbergMarquardt algorithm [25] offer state of the art performance on the BAL data set [1], with 3-5x reduction in execution time over currently available methods while delivering comparable or better solution quality.

show abstract

“…Each vector is represented as a pair of arrays of same length NNZ (number of nonzero entries); one of integer type for feature/target indices and the other of float type for values (float for possible decaying), with a same array index coupling an (index, value) pair. This data structure can be viewed as a flattened adaptation of Yale sparse matrix representation [10]. We further split the representations of features and targets for fast access of both, particularly the tensor product yi ⊗ xi, the major computational cost for model updates.…”

Section: Feature Vector Generation In O(1n)mentioning

confidence: 99%

Large-scale behavioral targeting

Chen

Pavlov

Canny

2009

Proceedings of the 15th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining

158

View full text Add to dashboard Cite

Behavioral targeting (BT) leverages historical user behavior to select the ads most relevant to users to display. The state-of-the-art of BT derives a linear Poisson regression model from fine-grained user behavioral data and predicts click-through rate (CTR) from user history. We designed and implemented a highly scalable and efficient solution to BT using Hadoop MapReduce framework. With our parallel algorithm and the resulting system, we can build above 450 BT-category models from the entire Yahoo's user base within one day, the scale that one can not even imagine with prior systems. Moreover, our approach has yielded 20% CTR lift over the existing production system by leveraging the well-grounded probabilistic model fitted from a much larger training dataset.Specifically, our major contributions include: (1) A MapReduce statistical learning algorithm and implementation that achieve optimal data parallelism, task parallelism, and load balance in spite of the typically skewed distribution of domain data. (2) An in-place feature vector generation algorithm with linear time complexity O(n) regardless of the granularity of sliding target window. (3) An in-memory caching scheme that significantly reduces the number of disk IOs to make large-scale learning practical. (4) Highly efficient data structures and sparse representations of models and data to enable fast model updates. We believe that our work makes significant contributions to solving large-scale machine learning problems of industrial relevance in general. Finally, we report comprehensive experimental results, using industrial proprietary codebase and datasets.

show abstract

A Comparison of Several Bandwidth and Profile Reduction Algorithms

Cited by 87 publications

References 2 publications

Direct-numerical simulation of the glottal jet and vocal-fold dynamics in a three-dimensional laryngeal model

Direct-numerical simulation of the glottal jet and vocal-fold dynamics in a three-dimensional laryngeal model

Visibility Based Preconditioning for bundle adjustment

Large-scale behavioral targeting

Contact Info

Product

Resources

About