query
stringlengths 1
13.4k
| pos
stringlengths 1
61k
| neg
stringlengths 1
63.9k
| query_lang
stringclasses 147
values | __index_level_0__
int64 0
3.11M
|
|---|---|---|---|---|
Thresholding-based Iterative Selection Procedures for Model Selection and Shrinkage | Relaxed lasso | An overset-grid method for 3D unsteady incompressible flows | eng_Latn | 800 |
Closest pair and the post office problem for stochastic points | An optimal algorithm for approximate nearest neighbor searching fixed dimensions | Characterization of glucose-sensitive insulin release systems in simulated in vivo conditions | eng_Latn | 801 |
Sparse inverse covariance estimation with the graphical lasso | Pathwise coordinate optimization | Computer simulation of the role of protein corona in cellular delivery of nanoparticles. | eng_Latn | 802 |
Matrix Completion Under Monotonic Single Index Models | Low-rank matrix completion by riemannian optimization | Low-loss and compact 2.4-GHz CMOS bandpass filter with finite transmission zeros | zsm_Latn | 803 |
LSRN: A Parallel Iterative Solver for Strongly Over- or Under-Determined Systems | Finding structure with randomness: Probabilistic algorithms for constructing approximate matrix decompositions | Photo tourism: exploring photo collections in 3D | eng_Latn | 804 |
Diagonal scaling in Douglas-Rachford splitting and ADMM | a fast iterative shrinkage - thresholding algorithm for linear inverse problems . | GoDec: Randomized Lowrank & Sparse Matrix Decomposition in Noisy Case | eng_Latn | 805 |
an interior - point gradient method for large - scale totally nonnegative least squares problems ∗ . | Algorithms for Non-negative Matrix Factorization | Maximum Likelihood Reconstruction for Emission Tomography | eng_Latn | 806 |
Robust Subspace Segmentation with Block-Diagonal Prior | Efficient subspace segmentation via quadratic programming | Security and Protocol Exploit Analysis of the 5G Specifications | eng_Latn | 807 |
Sequential Quadratic Programming (SQP) for optimal control in direct numerical simulation of turbulent flow | A Globally Convergent Filter Method for Nonlinear Programming | Web table column categorisation and profiling | eng_Latn | 808 |
A fast randomized algorithm for orthogonal projection | numerical methods for least square problems . | Approximate nearest neighbors and the fast Johnson-Lindenstrauss transform | eng_Latn | 809 |
Improved Approximation Algorithms for Large Matrices via Random Projections | Approximate nearest neighbors and the fast Johnson-Lindenstrauss transform | Glycosylation in cancer: mechanisms and clinical implications | eng_Latn | 810 |
Approximate Gaussian Elimination | Approximate Gaussian Elimination for Laplacians - Fast, Sparse, and Simple | path finding methods for linear programming : solving linear programs in o ( vrank ) iterations and faster algorithms for maximum flow . | eng_Latn | 811 |
An Efficient Globally Optimal Algorithm for Asymmetric Point Matching. | A shortest augmenting path algorithm for dense and sparse linear assignment problems | Neural network models for supporting drug and multidrug resistant tuberculosis screening diagnosis | eng_Latn | 812 |
Preconditioned temporal difference learning | Iterative Methods for Sparse Linear Systems | The Convergence of TD(λ) for General λ | eng_Latn | 813 |
Fast Robust PCA on Graphs | proximal splitting methods in signal processing ∗ . | Drug-induced renal Fanconi syndrome | kor_Hang | 814 |
An Empirical Study of ADMM for Nonconvex Problems | A dual algorithm for the solution of nonlinear variational problems via finite element approximation | Scaling Multicore Databases via Constrained Parallel Execution | kor_Hang | 815 |
A PROXIMAL POINT ALGORITHM FOR LOG-DETERMINANT OPTIMIZATION WITH GROUP LASSO REGULARIZATION | Robust uncertainty principles: exact signal reconstruction from highly incomplete frequency information | i FEM : AN INNOVATIVE FINITE ELEMENT METHOD PACKAGE IN MATLAB | yue_Hant | 816 |
Modeling disease progression via multi-task learning | LSQR: An Algorithm for Sparse Linear Equations and Sparse Least Squares | Statistical parametric maps in functional imaging: a general linear approach | eng_Latn | 817 |
GPOPS-II: A MATLAB Software for Solving Multiple-Phase Optimal Control Problems Using hp-Adaptive Gaussian Quadrature Collocation Methods and Sparse Nonlinear Programming | On the implementation of an interior-point filter line-search algorithm for large-scale nonlinear programming | Integrating Additional Chord Information Into HMM-Based Lyrics-to-Audio Alignment | eng_Latn | 818 |
Non-local sparse and low-rank regularization for structure-preserving image smoothing | Robust Principal Component Analysis: Exact Recovery of Corrupted Low-Rank Matrices via Convex Optimization | Traumatic perforation of the hypopharynx--an unusual form of abuse. | eng_Latn | 819 |
Fast Gaussian Process Regression using KD-Trees | Iterative Methods for Sparse Linear Systems | Effective route planning in road networks using multi constraint routing algorithm | eng_Latn | 820 |
Sparse Manifold Clustering and Embedding | A Global Geometric Framework for Nonlinear Dimensionality Reduction | Fast Measurement Technique for Phased Array Calibration | eng_Latn | 821 |
Multi-channel Weighted Nuclear Norm Minimization for Real Color Image Denoising | computing the nearest correlation matrix — a problem from finance ∗ . | Radar Cross Section Reduction of a Microstrip Antenna Based on Polarization Conversion Metamaterial | eng_Latn | 822 |
Partial Sum Minimization of Singular Values in Robust PCA: Algorithm and Applications | Guaranteed Rank Minimization via Singular Value Projection | An accelerated gradient method for trace norm minimization | eng_Latn | 823 |
dictionary learning . | A Global Geometric Framework for Nonlinear Dimensionality Reduction | control techniques in heating , ventilating and air conditioning ( hvac ) systems . | eng_Latn | 824 |
Blind Deconvolution Using Convex Programming | A Simpler Approach to Matrix Completion | Guaranteed Minimum-Rank Solutions of Linear Matrix Equations via Nuclear Norm Minimization | eng_Latn | 825 |
Robust volume minimization-based matrix factorization via alternating optimization | A Variable Splitting Augmented Lagrangian Approach to Linear Spectral Unmixing | Deformation embedding for point-based elastoplastic simulation | eng_Latn | 826 |
A Deterministic Analysis for LRR | Robust Recovery of Subspace Structures by Low-Rank Representation | Aspergillus fumigatus and related species. | kor_Hang | 827 |
Unconstrained inverse quadratic programming problem | Inverse Combinatorial Optimization: A Survey on Problems, Methods, and Results | Least-squares orthogonal distances fitting of circle, sphere, ellipse, hyperbola, and parabola | eng_Latn | 828 |
Tighter Linear Program Relaxations for High Order Graphical Models | SLIC Superpixels Compared to State-of-the-Art Superpixel Methods | Lazy snapping | eng_Latn | 829 |
Robust Statistical Face Frontalization | A Singular Value Thresholding Algorithm for Matrix Completion | Robot Path Integration in Manufacturing Processes: Genetic Algorithm Versus Ant Colony Optimization | eng_Latn | 830 |
CVXGEN : a code generator for embedded convex optimization | SOLUTION OF SPARSE RECTANGULAR SYSTEMS USING LSQR AND CRAIG | The Impact of Small Business Enterprises on the Economy of Trinidad & Tobago | eng_Latn | 831 |
Accurate Non-Iterative O(n) Solution to the PnP Problem | Complete solution classification for the perspective-three-point problem | The cup runneth over: lessons from the ever-expanding pool of primary immunodeficiency diseases | eng_Latn | 832 |
Secure matrix generation for Compressive Sensing embedded cryptography | Real-time compressive tracking | Multiscale Molecular Simulations of Polymer-Matrix Nanocomposites | eng_Latn | 833 |
Sparse inverse covariance estimation with the graphical lasso | Pathwise coordinate optimization | Reflection Phase Characterization of Curved High Impedance Surfaces | eng_Latn | 834 |
A Partial Derandomization of PhaseLift using Spherical Designs | Recovering low-rank matrices from few coefficients in any basis | Guaranteed Minimum-Rank Solutions of Linear Matrix Equations via Nuclear Norm Minimization | eng_Latn | 835 |
Unified representation of tractography and diffusion-weighted MRI data using sparse multidimensional arrays | Higher Order Partial Least Squares (HOPLS): A Generalized Multilinear Regression Method | nonnegative matrix and tensor factorizations . | eng_Latn | 836 |
New support vector algorithms with parametric insensitive/margin model | Shrinking the Tube: A New Support Vector Regression Algorithm | Cell-laden microengineered gelatin methacrylate hydrogels | eng_Latn | 837 |
Robust high-dimensional precision matrix estimation | Sparse inverse covariance estimation with the graphical lasso | Analysis and Synthesis of Geneva Mechanism with Elliptic Crank | eng_Latn | 838 |
Removing Clouds and Recovering Ground Observations in Satellite Image Sequences via Temporally Contiguous Robust Matrix Completion | Spectral regularization algorithms for learning large incomplete matrices | Effects of dihydrocapsiate on adaptive and diet-induced thermogenesis with a high protein very low calorie diet: a randomized control trial | eng_Latn | 839 |
Categorical matrix completion | Exact Matrix Completion via Convex Optimization | CORAL-SDN: A software-defined networking solution for the Internet of Things | eng_Latn | 840 |
Broadcast-based distributed alternating direction method of multipliers | Distributed Sparse Linear Regression | understanding the empirical hardness of np - complete problems . | eng_Latn | 841 |
Hyperspectral Unmixing Overview: Geometrical, Statistical, and Sparse Regression-Based Approaches | An overview on hyperspectral unmixing: Geometrical, statistical, and sparse regression based approaches | Geodesic Convolutional Shape Optimization | kor_Hang | 842 |
Analysis and Design of Optimization Algorithms via Integral Quadratic Constraints | a fast iterative shrinkage - thresholding algorithm for linear inverse problems . | Power signatures of high-performance computing workloads | eng_Latn | 843 |
Fast Stochastic Algorithms for SVD and PCA: Convergence Properties and Convexity | Matrix analysis | sensorimotor neuropathy in a patient with marinesco - sjogren syndrome . | eng_Latn | 844 |
Fast Bayesian Matching Pursuit: Model Uncertainty and Parameter Estimation for Sparse Linear Models | Regression Shrinkage and Selection Via the Lasso | Generation and validation of virtual point cloud data for automated driving systems | eng_Latn | 845 |
Sparse Unmixing of Hyperspectral Data | Vertex component analysis: a fast algorithm to unmix hyperspectral data | Two Is Bigger (and Better) Than One: the Wikipedia Bitaxonomy Project | kor_Hang | 846 |
Distributed Optimization with Arbitrary Local Solvers | Statistical learning theory | Innervation of periosteum and bone by sympathetic vasoactive intestinal peptide-containing nerve fibers. | eng_Latn | 847 |
The Mutual Information in Random Linear Estimation | Mutual information for symmetric rank-one matrix estimation: A proof of the replica formula | An Enhanced Computer Simulation Model of the IEEE Std C62.41.2-2002 Surge Generator for Simulated Surge Testing of Electrical Systems | eng_Latn | 848 |
An SVD-free Pareto curve approach to rank minimization | Image decomposition via the combination of sparse representations and a variational approach | Low Self-Esteem Is Related to Aggression, Antisocial Behavior, and Delinquency | eng_Latn | 849 |
A Parallel Sparse Direct Solver via Hierarchical DAG Scheduling | Highly Scalable Parallel Algorithms for Sparse Matrix Factorization | Parametric Localization of Di&ributed 'Sources | eng_Latn | 850 |
A path following algorithm for Sparse Pseudo-Likelihood Inverse Covariance Estimation (SPLICE) | Regression Shrinkage and Selection Via the Lasso | A Direct Least-Squares Solution to the PnP Problem with Unknown Focal Length | kor_Hang | 851 |
Working Locally Thinking Globally - Part I: Theoretical Guarantees for Convolutional Sparse Coding | orthogonal least squares methods and their application to non - linear system identification . | A modification of Shanks' baby-step giant-step algorithm | eng_Latn | 852 |
An efficient ADMM algorithm for multidimensional anisotropic total variation regularization problems | On the Douglas-Rachford splitting method and the proximal point algorithm for maximal monotone operators | End-to-End Reinforcement Learning for Automatic Taxonomy Induction | eng_Latn | 853 |
Spacecraft telemetry data monitoring by dimensionality reduction techniques | Laplacian Eigenmaps for Dimensionality Reduction and Data Representation | PHASE LOCKED LOOPS-I 1 ECE-599 : Low Power , Adaptive Bandwidth Tracking Phase Locked Loop Design | eng_Latn | 854 |
An Optimal Algorithm for Online Square Detection | A square is the concatenation of two identical non-empty strings. Let S be the input string which is given character by character. Let m be the (unknown) smallest integer such that the m-th prefix of S contains a square. The online square detection problem is to determine m as soon as the m-th character of S is read. The best previously known algorithm of the online square detection problem, due to Leung, Peng, and Ting, runs in O(mlog2m) time. We improve the time complexity to O(mlog β), where β is the number of distinct characters in the m-th prefix of the input string. It is not difficult to implement our algorithm to run in expected O(m) time. | In this paper, we provide theoretical analysis for a cubic regularization of Newton method as applied to unconstrained minimization problem. For this scheme, we prove general local convergence results. However, the main contribution of the paper is related to global worst-case complexity bounds for different problem classes including some nonconvex cases. It is shown that the search direction can be computed by standard linear algebra technique. | kor_Hang | 855 |
Fundamental solutions for the collation method in planar elastostatics | A method for solving planar elastostatic problems referred to as the fundamental collocation method is described. The method features techniques employed in the boundary integral equation method and the boundary-point-least-squares collocation method. The governing equations are satisfied using fundamental solutions of the theory of elasticity corresponding to concentrated forces while boundary conditions are satisfied using the BPLS collaction technique. Numerical results are given for two sample problems. The method exhibits some favourable characteristics in comparison with other numerical methods. | Many data analysis problems involve an investigation of relationships between attributes in heterogeneous databases, where different prediction models can be more appropriate for different regions. We propose a technique of integrating global and local random subspace ensemble. We performed a comparison with other well known combining methods on standard benchmark datasets and the proposed technique gave better accuracy. | eng_Latn | 856 |
On stable perturbations for outer inverses of linear operators in Banach spaces | Abstract In this paper, we investigate stable perturbations and their characterizations for various types of outer inverses, such as generalized, { 2 , 3 } -, { 2 , 4 } -, { 2 , 5 } -, { 1 , 2 , 3 } -, { 1 , 2 , 4 } -, Moore–Penrose, group, Drazin and generalized Drazin inverses. Some known results are improved and extended. | Abstract Recently new optimal Krylov subspace methods have been discovered for normal matrices. In light of this, novel ways to quantify nonnormality are considered in connection with various families of matrices. We use as a criterion how, for a given matrix, these iterative methods introduced can be employed via, e.g., inexpensive matrix factorizations. The unitary orbit of the set of binormal matrices provides a natural extension of normal matrices. Its elements yield polynomially normal matrices of moderate degree. In this context several matrix nearness problems arise. | eng_Latn | 857 |
Fast projections onto mixed-norm balls with applications | Joint sparsity offers powerful structural cues for feature selection, especially for variables that are expected to demonstrate a "grouped" behavior. Such behavior is commonly modeled via group-lasso, multitask lasso, and related methods where feature selection is effected via mixed-norms. Several mixed-norm based sparse models have received substantial attention, and for some cases efficient algorithms are also available. Surprisingly, several constrained sparse models seem to be lacking scalable algorithms. We address this deficiency by presenting batch and online (stochastic-gradient) optimization methods, both of which rely on efficient projections onto mixed-norm balls. We illustrate our methods by applying them to the multitask lasso. We conclude by mentioning some open problems. | Consider arbitrary collections A = a_1,a_2,.. .,a_n of items and Q = q_1,q_2,...,q_m (1 leqslant mn leqslant n) of queries from a totally ordered universe. The multiple rank problem involves computing for every query qi the number of items in A that have a lesser value. Our contribution is to show that the problem at hand can be solved time-optimally on meshes with multiple broadcasting. More specifically, if the collection A is siored in some order one item per processor and if Q is stored one query per processor in the leftmost frac{m} {{sqrt n }} columns of a mesh with multiple broadcasting of size sqrt n x /sqrt n, the corresponding instance of the multiple rank problem can be solved in Theta left( {m^{frac{1} {3}} n^{frac{1} {6}} } right) time. As an application we present a time-optimal algorithm to compute the histogram of a m-level gray image of size sqrt n x sqrt n in Theta left( {m^{frac{1} {3}} n^{frac{1} {6}} } right) time. | eng_Latn | 858 |
Safety and efficacy of a novel algorithm to guide decision-making in high-risk interventional coronary procedures | Abstract Background Patients with severe coronary artery disease (CAD), comorbidities, or impaired hemodynamics are at risk during percutaneous coronary interventions. The aim of the study was to investigate the safety and efficacy of a novel risk-stratification algorithm for high-risk coronary procedures. Methods and results We prospectively screened 1189 patients with CAD requiring revascularization (period 07/2017–06/2018). The algorithm was designed to select high-risk procedures. Patients with elevated risk (n = 150) were classified into 3 risk groups (high-risk intervention [HRI] I–III) and procedural management was adjusted according to HRI group. Overall, 55 patients were categorized as HRI I, 52 as HRI II, and 43 as HRI III. With increasing HRI-level, SYNTAX score increased (HRI I:15 ± 5% vs. HRI II:24 ± 8% vs. HRI III:34 ± 7%; p Conclusions The novel algorithm is a safe team-based stratification method for the identification and management of patients undergoing high-risk coronary interventions. | In this paper, we present a large-update primal-dual interior-point method for symmetric cone optimization (SCO) based on a new kernel function, which determines both search directions and the proximity measure between the iterate and the center path. The kernel function is neither a self-regular function nor the usual logarithmic kernel function. Besides, by using Euclidean Jordan algebraic techniques, we achieve the favorable iteration complexity \(O(\sqrt r (\log r)^2 \log (r/\varepsilon ))\), which is as good as the convex quadratic semi-definite optimization analogue. | eng_Latn | 859 |
The Existence of Three Positive Solutions for Second-order Multi-point Boundary Value Problems in Banach Spaces | By using fixed point theorems of strict set contraction mapping,the boundary value problems for compact type conditions in Banach spaces are discussed.First,a fixed point theorem is extended to strict set contraction mapping.And then the suitable functionals are constructed to prove the existence of three positive solutions for second-order n-point boundary value problem in Banach spaces by using the fixed point theorem proved above.At last,an example is given to demonstrate the availablity of the result. | Abstract Recently new optimal Krylov subspace methods have been discovered for normal matrices. In light of this, novel ways to quantify nonnormality are considered in connection with various families of matrices. We use as a criterion how, for a given matrix, these iterative methods introduced can be employed via, e.g., inexpensive matrix factorizations. The unitary orbit of the set of binormal matrices provides a natural extension of normal matrices. Its elements yield polynomially normal matrices of moderate degree. In this context several matrix nearness problems arise. | eng_Latn | 860 |
Preconditioned conjugate gradient method for the non‐linear finite element analysis with particular reference to 3D reinforced concrete structures | A hierarchically preconditioned conjugate gradient (PCG) method for finite element analysis is presented. Its use is demonstrated for the difficult problem of the non‐linear analysis of 3D reinforced concrete structures. Examples highlight the dramatic savings in computer storage and more modest savings in solution times obtained using PCG especially for large problems. | This paper first studies the preference information represented by means of preference ordering on group decision-making problems.By using the definition 2.1,prefererce ordering preference in formation of each decision maker can be transformed into his/her complementary judgement matrices .In this paper,it is prover that the complementary judgement matrices are fuzzy consistency judgement matrices.The normalizing rank aggregation method of group complementary judgement matrices is utilized to order alternatives.Finally,an algorithm example is given. | eng_Latn | 861 |
MEROMOR0PHIC UNIVALENT HARMONIC FUNCTIONS WITH NEGATIVE COEFFICIENTS | The purpose of this paper is to give sufficient coefficient conditions for a class of univalent harmonic functions that map each $z$ = r >1 onto a curve that bounds a domain that is starlike with respect to origin. Furthermore, it is shown that these conditions are also necessary when the coefficients are negative. Extreme points for these classes are also determined. Finally, comparable results are given for the convex analgo. | Abstract Recently new optimal Krylov subspace methods have been discovered for normal matrices. In light of this, novel ways to quantify nonnormality are considered in connection with various families of matrices. We use as a criterion how, for a given matrix, these iterative methods introduced can be employed via, e.g., inexpensive matrix factorizations. The unitary orbit of the set of binormal matrices provides a natural extension of normal matrices. Its elements yield polynomially normal matrices of moderate degree. In this context several matrix nearness problems arise. | yue_Hant | 862 |
Fault tolerant matrix triangularization and solution of linear systems of equations | The authors present a fault tolerant algorithm for the solution of linear systems of equations using matrix triangularization procedures suitable for implementation on array architectures. Gaussian elimination with partial or pairwise pivoting and QR decomposition are made fault tolerant against two transient errors occurring during the triangularization procedure. The extended Euclidean algorithm is implemented to solve for the locations and values of the errors defined appropriately using the theory of error correcting codes. The Sherman-Morrison Woodbury formula is then used to obtain the correct solution vector to the linear system of equations without requiring a valid decomposition. > | We consider the problem of finding the cheapest routing for a set of commodities over a directed graph, such that: i) each commodity flows through a single path, ii) the routing cost of each arc is given by a convex piecewise linear function of the load (i.e. the total flow) traversing it. We propose a new mixed-integer programming formulation for this problem. This formulation gives a complete description of the associated polyhedron for the single commodity case, and produces very tight linear programming bounds for the multi-commodity case. | eng_Latn | 863 |
Block Preconditioners for Saddle Point Problems | A general purpose block LU preconditioner for saddle point problems is presented. A major difference between the approach presented here and that of other studies is that an explicit, accurate approximation of the Schur complement matrix is efficiently computed. This is used to obtain a preconditioner to the Schur complement matrix which in turn defines a preconditioner for the global system. A number of variants are developed and results are reported for a few linear systems arising from CFD applications. | We consider a stable driven degenerate stochastic differential equation, whose coefficients satisfy a kind of weak H{\"o}rmander condition. Under mild smoothness assumptions we prove the uniqueness of the martingale problem for the associated generator under some dimension constraints. Also, when the driving noise is scalar and tempered, we establish density bounds reflecting the multi-scale behavior of the process. | eng_Latn | 864 |
Matrix Routh-approximant reduced-order modelling for multivariable systems | A matrix Routh-approximant modelling procedure is proposed for a multi-input multi-output system characterized by a matrix transfer function G(s), where G(s) = B(s) A −1(s) and A(s) and B(s) are matrix polynomials in s. The associated time-domain modelling procedure is also discussed. Compared with three matrix Cauer continued-fractions, the proposed method requires less computational effort. In addition, it is applicable to systems with unequal number of inputs and outputs. A demonstrative example is included. | This paper details the development of a simplified MRP through the equation of Leontief (linear algebra), and proposes its teaching in three parts a) data collection and recognition of the relationship between materials, b) a list of variables and mathematics explanation, c) obtaining results and verifying them by software. The academic exercise is about the flow of ingredients required to comply with a plan for food services and is intended to be a learning tool to facilitate the study of this method in the hospitality management courses. | eng_Latn | 865 |
Accelerated diagonal gradient-type method for large-scale unconstrained optimization | In this study, we propose an accelerated diagonal-updating scheme for solving large-scale optimization, where a scaled diagonal matrix is used to approximate the Hessian. We combine an accelerator with the diagonal-updating method to improve the efficiency of the algorithm. This accelerator is employed to ensure that the function value can be reduced significantly at each step. Moreover, the algorithm employs a suitable monotone strategy to guarantee the global convergence of the algorithm. Several numerical results are reported, which demonstrate that the proposed method is promising and more robust than other diagonal updating schemes. | Jointly Gaussian memoryless sources are observed at N distinct terminals. The goal is to efficiently encode the observations in a distributed fashion so as to enable reconstruction of any one of the observations, say the first one, at the decoder subject to a quadratic fidelity criterion. Our main result is a precise characterization of the rate-distortion region when the covariance matrix of the sources satisfies a ?tree-structure? condition. In this situation, a natural analog-digital separation scheme optimally trades off the distributed quantization rate tuples and the distortion in the reconstruction: each encoder consists of a point-to-point Gaussian vector quantizer followed by a Slepian-Wolf binning encoder. We also provide a partial converse that suggests that the tree-structure condition is fundamental. | eng_Latn | 866 |
Froissart doublets in Padé approximation in the case of polynomial noise | First, we study the relation between the zeros of random polynomials Rn+1 and the zeros and poles of their Pade approximants [n/n]Rn+1. Next, we consider the distribution of zeros and poles of Pade approximants to the geometric series perturbed by a random polynomial noise. We observe numerically interesting connections between two above problems. Some numerical observations on the Froissart doublets have been also made. | We analyze the parallelization of QR factorization by means of Householder transformations. This parallelization is carried out on a machine with a mesh topology (a 2-D torus to be more precise). We use a cyclic distribution of the elements of the sparse matrix M we want to decompose over the processors. Each processor represents the nonzero elements of its part of the matrix by a one-dimensional doubly linked list data structure. Then, we describe the different procedures that constitute the parallel algorithm. As an application of QR factorization, we concentrate on the least squares problem and finally we present an evaluation of the efficiency of this algorithm for a set of test matrices from the Harwell-Boeing sparse matrix collection. | eng_Latn | 867 |
Nicoleta Breaz-The cross-validation method in the smoothing spline regression THE CROSS-VALIDATION METHOD IN THE SMOOTHING SPLINE REGRESSION | Abstract. One of the goals, in the context of nonparametric regression by smoothing spline functions, is to choose the optimal value for the smoothing parameter. In this paper, we deal with the cross validation method(CV), as a performance criteria for smoothing parameter selection. First, we implement a CV-based algorithm, in Matlab 6.5 medium and we apply it on a test function, in order to emphase the quality of the fitting by the CV-smoothing spline function. Then, we fit some real data with this kind of function. | Principal skewness analysis (PSA) has been introduced to the remote sensing community recently, which is equivalent to fast independent component analysis (FastICA) when skewness is considered as a non-Gaussian index. However, similar to FastICA, PSA also has the nonconvergence problem in searching for optimal projection directions. In this letter, we propose a new iteration strategy to alleviate PSA's nonconvergence problem, and we name this new version of PSA as momentum PSA (MPSA). MPSA still adopts the same fixed-point algorithm as PSA does. Different from PSA, the $(k+1)$ th result in the iteration process of MPSA not only depends on the $k$ th iteration result but also is related to the $(k-1)$ th iteration. Experiments conducted for both simulated data and real-world hyperspectral image demonstrate that MPSA has an obvious advantage over PSA in convergence performance and computational speed. | yue_Hant | 868 |
A globally optimal tri-vector method to solve an ill-posed linear system | In the present paper, a future cone in the Minkowski space defined in terms of the square-norm of the residual vector for an ill-posed linear system to be solved, is used to derive a nonlinear system of ordinary differential equations. Then the forward Euler scheme is used to generate an iterative algorithm. Two critical values in the critical descent tri-vector are derived, which lead to the largest convergence rate of the resultant iterative algorithm, namely the globally optimal tri-vector method (GOTVM). Some numerical examples are used to reveal the superior performance of the GOTVM than the famous methods of conjugate gradient (CGM) and generalized minimal residual (GMRES). Through the numerical tests we also set forth the rationale by assuming the tri-vector as being a better descent direction. | We have examined the linear stability of triangular equilibrium points in restricted three body problem. We supposed bigger primary as an oblate spheroid and other primary as source of radiation. We have found characteristic equation of the problem. All the four roots of characteristic equation are pure imaginary. Hence we conclude that triangular equilibrium points are stable. | eng_Latn | 869 |
Fast Robust Model Selection in Large Datasets | Large datasets are increasingly common in many research fields. In particular, in the linear regression context, it is often the case that a huge number of potential covariates are available to explain a response variable, and the first step of a reasonable statistical analysis is to reduce the number of covariates. This can be done in a forward selection procedure that includes selecting the variable to enter, deciding to retain it or stop the selection, and estimating the augmented model. Least squares plus t tests can be fast, but the outcome of a forward selection might be suboptimal when there are outliers. In this article we propose a complete algorithm for fast robust model selection, including considerations for huge sample sizes. Because simply replacing the classical statistical criteria with robust ones is not computationally possible, we develop simplified robust estimators, selection criteria, and testing procedures for linear regression. The robust estimator is a one-step weighted M-estimato... | Consider arbitrary collections A = a_1,a_2,.. .,a_n of items and Q = q_1,q_2,...,q_m (1 leqslant mn leqslant n) of queries from a totally ordered universe. The multiple rank problem involves computing for every query qi the number of items in A that have a lesser value. Our contribution is to show that the problem at hand can be solved time-optimally on meshes with multiple broadcasting. More specifically, if the collection A is siored in some order one item per processor and if Q is stored one query per processor in the leftmost frac{m} {{sqrt n }} columns of a mesh with multiple broadcasting of size sqrt n x /sqrt n, the corresponding instance of the multiple rank problem can be solved in Theta left( {m^{frac{1} {3}} n^{frac{1} {6}} } right) time. As an application we present a time-optimal algorithm to compute the histogram of a m-level gray image of size sqrt n x sqrt n in Theta left( {m^{frac{1} {3}} n^{frac{1} {6}} } right) time. | yue_Hant | 870 |
A safe technique for radical antegrade modular pancreatosplenectomy with venous resection for pancreatic cancer. | Received May 13, 2013; Revised August 9, 2013; Accepted A From Hepato-Pancreato-Biliary Surgery and Liver Transp des Pathologies Digestives, Hepatiques et de la Transplan de Hautepierre-Hopitaux Universitaires de Strasbourg, Uni bourg, Strasbourg, France. Correspondence address: Philippe Bachellier, MD, P Pancreato-Biliary Surgery and Liver Transplantation, Pole Digestives, Hepatiques et de la Transplantation, Hopital d Hopitaux Universitaires de Strasbourg, Universite de Strasbo Moliere, 67098, Strasbourg, France. email: Philip chru-strasbourg.fr | In the present paper, a future cone in the Minkowski space defined in terms of the square-norm of the residual vector for an ill-posed linear system to be solved, is used to derive a nonlinear system of ordinary differential equations. Then the forward Euler scheme is used to generate an iterative algorithm. Two critical values in the critical descent tri-vector are derived, which lead to the largest convergence rate of the resultant iterative algorithm, namely the globally optimal tri-vector method (GOTVM). Some numerical examples are used to reveal the superior performance of the GOTVM than the famous methods of conjugate gradient (CGM) and generalized minimal residual (GMRES). Through the numerical tests we also set forth the rationale by assuming the tri-vector as being a better descent direction. | eng_Latn | 871 |
An algebraic approach based on symmetry relations of state space matrices for decoupling the dynamics of a large space structure | In order to design an effective control policy for the 10 m diameter primary mirror of the Grantecan Telescope, the coupling among the segments that compose the mirror through the structure that supports it must be treated. We present the mathematical demonstration of two important results in order to deal with the mirror dynamics decoupling. Those are: 1) the multiplicity of the eigenvalue corresponding to the mirror segments decreases as the number of structure modes considered increases, and 2) the vectorial subspace composed of the eigenvectors corresponding to the structure and the segments coupled with it is orthogonal to the one composed of the eigenvectors corresponding to the rest of segments. | Many data analysis problems involve an investigation of relationships between attributes in heterogeneous databases, where different prediction models can be more appropriate for different regions. We propose a technique of integrating global and local random subspace ensemble. We performed a comparison with other well known combining methods on standard benchmark datasets and the proposed technique gave better accuracy. | eng_Latn | 872 |
Efficient Computation of the Binary Vector That Maximizes a Rank-Deficient Quadratic Form | The maximization of a full-rank quadratic form over a finite alphabet is NP-hard in both a worst-case sense and an average sense. Interestingly, if the rank of the form is not a function of the problem size, then it can be maximized in polynomial time. An algorithm for the efficient computation of the binary vector that maximizes a rank-deficient quadratic form is developed based on an analytic procedure. Auxiliary spherical coordinates are introduced and the multi-dimensional space is partitioned into a polynomial-size set of regions; each region corresponds to a distinct binary vector. The binary vector that maximizes the rank-deficient quadratic form is shown to belong to the polynomial-size set of candidate vectors. Thus, the size of the feasible set is efficiently reduced from exponential to polynomial. | Abstract A method, implementable on a digital computer, for fast writing of the transfer function matrix of a linear time invariant compartmental model in its symbolic expression is given. Theoretical fundamentals are proved, and a digital computer implementation of the procedure is given. An example is described. | eng_Latn | 873 |
Sparse Householder QR factorization on a mesh | We analyze the parallelization of QR factorization by means of Householder transformations. This parallelization is carried out on a machine with a mesh topology (a 2-D torus to be more precise). We use a cyclic distribution of the elements of the sparse matrix M we want to decompose over the processors. Each processor represents the nonzero elements of its part of the matrix by a one-dimensional doubly linked list data structure. Then, we describe the different procedures that constitute the parallel algorithm. As an application of QR factorization, we concentrate on the least squares problem and finally we present an evaluation of the efficiency of this algorithm for a set of test matrices from the Harwell-Boeing sparse matrix collection. | Bottom-up holographic models of QCD, inspired by the anti-de Sitter space/conformal field theory correspondence, have shown a remarkable degree of phenomenological success. However, they rely on a number of bold assumptions. We investigate the reliability of one of the key assumptions, which involves matching the parameters of these models to QCD at high 4D momentum q{sup 2} and renormalization scale {mu}{sup 2}. We show that this leads to phenomenological and theoretical inconsistencies for scale-dependent quantities such as . | eng_Latn | 874 |
Quarter-tree based on method for gray image compression | Based on the local statistical property of gray image, the compression method for gray image is investigated. The basic point is that, if the maxim difference of gray value of the current original sub-block is lower than a given threshold, the current original sub-block can be considered as an integral sub-block, otherwise the original sub-block is decomposed into four smaller original sub-block with the same size. In so doing the original image is decomposed into lots of integral sub-blocks, and the position, size and gray value of each integral sub-block can be stored with an efficient way, so as to achieve the aim of compression. The method of this paper uses reasonable smoothing algorithm, and removes block-effect satisfactorily. Computer simulation experiments show that the method can gain higher CR (compression-rate) and PSNR (peak-signal-noise-rate), keep details where pixel gray value varies hugely, remove block-effect conveniently, and have higher practical value. | We analyze the parallelization of QR factorization by means of Householder transformations. This parallelization is carried out on a machine with a mesh topology (a 2-D torus to be more precise). We use a cyclic distribution of the elements of the sparse matrix M we want to decompose over the processors. Each processor represents the nonzero elements of its part of the matrix by a one-dimensional doubly linked list data structure. Then, we describe the different procedures that constitute the parallel algorithm. As an application of QR factorization, we concentrate on the least squares problem and finally we present an evaluation of the efficiency of this algorithm for a set of test matrices from the Harwell-Boeing sparse matrix collection. | eng_Latn | 875 |
Aspects of nonnormality for iterative methods | Abstract Recently new optimal Krylov subspace methods have been discovered for normal matrices. In light of this, novel ways to quantify nonnormality are considered in connection with various families of matrices. We use as a criterion how, for a given matrix, these iterative methods introduced can be employed via, e.g., inexpensive matrix factorizations. The unitary orbit of the set of binormal matrices provides a natural extension of normal matrices. Its elements yield polynomially normal matrices of moderate degree. In this context several matrix nearness problems arise. | Application of the method group theory to structure and analyse 9 normal vibration modes.According to both infrared and Raman spectrum select formula,we can get the spectral characteristic of CH4.Group theory plays a important role in this paper,using it we can do our best to ease the calculation. | eng_Latn | 876 |
Positive solutions of elliptic equations in RN with a super-subcritical nonlinearity | We consider the problem: ::: ::: Δu+up+uq=0inRN,0<u(x)→0as|x|→+∞, ::: ::: where 1 NN−2 and then lets q approach N+2N−2. If q is fixed and p gets close enough to NN−2, then no solution exists. | We construct a diagnostic predictor for patient disease status based on a single data set of mass spectra of serum samples together with the binary case-control response. The model is logistic regression with Bernoulli log-likelihood augmented either by quadratic ridge or absolute L1 penalties. For ridge penalization using the singular value decomposition we reduce the number of variables for maximization to the rank of the design matrix. With log-likelihood loss, 10-fold cross-validatory choice is employed to specify the penalization hyperparameter. Predictive ability is judged on a set-aside subset of the data. | eng_Latn | 877 |
Real-time compression coding based on convolution fractal image | In fractal image compression, an image is coded as a set of contractive transformations, and is guaranteed to generate an approximation to the original image when iteration applied to any initial image. In this paper, according to Jacquin' 5 PIFS algorithm, and by analyzing traditional fractal mapping parameters, a kind of convolution-based fast fractal image coding scheme (CBFC) is advanced. To speed up the encoding and improve the compression ratio, it is combined with quad-tree partitioning neighbor searching algorithm. To improve the real-time performance of the algorithm, it is performed on TMS320C6201. Experiments results of algorithms based on CBFC, and CBFC using quad-tree partitioning structure on DSP are given in this paper as comparisons. The results show that fractal image real-time coding can be realized with the considerable reconstructive image and coding time. | We analyze the parallelization of QR factorization by means of Householder transformations. This parallelization is carried out on a machine with a mesh topology (a 2-D torus to be more precise). We use a cyclic distribution of the elements of the sparse matrix M we want to decompose over the processors. Each processor represents the nonzero elements of its part of the matrix by a one-dimensional doubly linked list data structure. Then, we describe the different procedures that constitute the parallel algorithm. As an application of QR factorization, we concentrate on the least squares problem and finally we present an evaluation of the efficiency of this algorithm for a set of test matrices from the Harwell-Boeing sparse matrix collection. | eng_Latn | 878 |
Real-time microgravimetric quantification of Cryptosporidium parvum in the presence of potential interferents | Abstract The quartz crystal microbalance with dissipation monitoring (QCM-D) is used to develop a biosensor for detection of viable Cryptosporidium parvum ( C. parvum ) in water matrices of varying complexity. In a clean environment, a good log–log linear response is obtained for detection of C. parvum in aqueous suspensions with oocyst concentrations from 3 × 10 5 to 1 × 10 7 oocysts/mL. C. parvum detection is slightly affected by the presence of dissolved organic acids, likely due to steric stabilization and/or masking of the antibodies/antigens by adsorbed molecules. Colloidal contaminants generally have a greater influence as biosensor interferents, whereby the presence of model latex microspheres, Enterococcus faecalis , or Escherichia coli , led to decreases in biosensor response of up to 64%, 40%, and 20%, respectively. | We analyze the parallelization of QR factorization by means of Householder transformations. This parallelization is carried out on a machine with a mesh topology (a 2-D torus to be more precise). We use a cyclic distribution of the elements of the sparse matrix M we want to decompose over the processors. Each processor represents the nonzero elements of its part of the matrix by a one-dimensional doubly linked list data structure. Then, we describe the different procedures that constitute the parallel algorithm. As an application of QR factorization, we concentrate on the least squares problem and finally we present an evaluation of the efficiency of this algorithm for a set of test matrices from the Harwell-Boeing sparse matrix collection. | eng_Latn | 879 |
[Kinins and oedema induced by different carrageenans (author's transl)]. | 1. The oedema induced in the paw of the Wistar rat by local injection of three different carrageenans allows us to determine the inflammatory activity of these sulphated polysaccharides. This activity is greatest for the lambda type, it is reduced for the iota compound and is smallest for the kappa carrageenan. 2. Reduction of plasmatic stores of kininogens by ellagic acid or inhibition of kinin formation by hexadimethrine reduce the inflammatory reaction induced by the three carrageenans. Inhibition of kininases by phenanthroline increases the oedema. Thus the kinin system is involved in the development of this type of oedema. 3. In Brown-Norway rats, which lack of plasmatic kallikrein and kininogens, the inflammatory activity of the three carrageenans is very small. By comparing the activity in Wistar rats and in Brown-Norway rats, it is concluded that the kinin system is the most important factor in the development of the inflammatory reaction induced by the three types of carrageenans. | We present an incremental approach to 2-norm estimation for triangular matrices. Our investigation covers both dense and sparse matrices which can arise for example from a QR, a Cholesky or a LU factorization. If the explicit inverse of a triangular factor is available, as in the case of an implicit version of the LU factorization, we can relate our results to incremental condition estimation (ICE). Incremental norm estimation (INE) extends directly from the dense to the sparse case without needing the modifications that are necessary for the sparse version of ICE. INE can be applied to complement ICE, since the product of the two estimates gives an estimate for the matrix condition number. Furthermore, when applied to matrix inverses, INE can be used as the basis of a rank-revealing factorization. | eng_Latn | 880 |
Multiplicative Watermark Detection Using Locally Optimum Nonlinearity | This paper concentrates on watermark detection problem, which judges the existence of watermark. A multiplicative watermark detector using locally optimum nonlinearity is derived, based on the condition that DWT coefficients obey to a generalized Gaussian distribution. Experimental results show that the suggested detector has good performances under various situations. Finally, a method for improving its responses is suggested. | We discuss the derivation of Markovian master equation via Nakajima-Zwanzig's projection operator method, when there exists initial correlation between the system and the reservoir. | eng_Latn | 881 |
Swa2p-dependent clathrin dynamics is critical for Flo11p processing and ‘Mat’ formation in the yeastSaccharomyces cerevisiae | The yeast Saccharomyces cerevisiae is able to form complex multicellular structures called mats on low-density agar Petri plates. Mat formation strictly depends on Flo11p, a cell surface mannoprotein that mediates the adhesion of yeast cells to the agar surface. Here, we show that Swa2p, an auxilin ortholog required for clathrin-coated vesicle uncoating, is strictly required for biofilm formation. We show that the maturation and cellular levels of Flo11p are affected in Deltaswa2 cells, yet without compromising invasive growth. Both the TPR and J-domains of Swa2p, but not its clathrin-binding and ubiquitin-association motifs, are required for its function in Flo11p processing. | Automated understanding of spatio-temporal usage patterns of real-world applications are significant in urban planning. With the capability of smartphones collecting various information using inbuilt sensors, the smart city data is enriched with multiple contexts. Whilst tensor factorization has been successfully used to capture latent factors (patterns) exhibited by the real-world datasets, the multifaceted nature of smart city data needs an improved modeling to utilize multiple contexts in sparse condition. Thus, in our ongoing research, we aim to model this data with a novel Context-Aware Nonnegative Coupled Sparse Matrix Tensor (CAN-CSMT) framework which imposes sparsity constraint to learn the true factors in sparse data. We also aim to develop a fast and efficient factorization algorithm to deal with the scalability problem persistent in the state-of-the-art factorization algorithms. | eng_Latn | 882 |
Updating the principal angle decomposition | A class of fast Householder-based sequential algorithms for updating the Principal Angle Decomposition is introduced. The updated Principal Angle Decomposition is of key importance in the adaptive implementation of several fundamental operations on correlated processes, such as adaptive Wiener filtering, rank-adaptive system identification, and rank and data compression concepts using canonical coordinates. An instructive example of rank-adaptive system identification is examined experimentally. | From the angles of appreciating the modern dance and the principle of modern dance movements,the paper elaborates on some of he experience in teaching modern dance. | eng_Latn | 883 |
Univocal Varimax: an Orthogonal Factor Rotation Program for Optimal Simple Structure | Univocal varimax is an orthogonal factor rotation strategy aimed at improving upon the simple structure qualities of a preliminary varimax solution. This is accomplished by targetting for patterned rotation the highest element in each row of the varimax factor loading matrix. This tends to yield a solution in which each variable in the final rotated matrix maximally loads on only one factor. Univocal varimax is particularly relevant to research problems in which each rotated factor should be marked by a relatively tight cluster of variables. A FORTRAN IV program is described for the efficient analysis of large input factor loading matrices. | For a general class of mixed models which includes the (Γ,γ)-model introduced by Shiryaev and Spokoiny (1993) we prove the minimaxity of a Pitman type estimator. This minimaxity is closely related to the asymptotic minimaxity of a sequence of Bayes estimators which is a consequence of the asymptotic shift invariance of the priors. Such priors are called asymptotically uniformly distributed | eng_Latn | 884 |
Graphene networks for high-performance flexible and transparent supercapacitors | Graphene network (GN) was synthesized by a two-step chemical vapour deposition (CVD) method, involving the thermal annealing sputter-coated Cu film to form a Cu network by annealing for CVD deposition of graphene onto the Cu network catalyst. The resultant graphene network was transferred onto a flexible and transparent polymer (e.g., PDMS) substrate while maintaining its porous structure and integrated interconnection, providing both good optical transparency (e.g., transmittance of 86% at 550 nm wavelength) and mechanical flexibility. Flexible and transparent all-solid-state supercapacitors based on the newly-developed graphene network were fabricated to exhibit an area specific capacitance of 4.2 μF cm−2 at a discharge current of 0.1 μA with a high optical transparency (transmittance of 84%), which outperforms devices based on uniform multi-layer graphene sheet. | Automated understanding of spatio-temporal usage patterns of real-world applications are significant in urban planning. With the capability of smartphones collecting various information using inbuilt sensors, the smart city data is enriched with multiple contexts. Whilst tensor factorization has been successfully used to capture latent factors (patterns) exhibited by the real-world datasets, the multifaceted nature of smart city data needs an improved modeling to utilize multiple contexts in sparse condition. Thus, in our ongoing research, we aim to model this data with a novel Context-Aware Nonnegative Coupled Sparse Matrix Tensor (CAN-CSMT) framework which imposes sparsity constraint to learn the true factors in sparse data. We also aim to develop a fast and efficient factorization algorithm to deal with the scalability problem persistent in the state-of-the-art factorization algorithms. | eng_Latn | 885 |
Effect of single-electron interface trapping in decanano MOSFETs: A 3D atomistic simulation study | Abstract We study the effect of trapping/detrapping of a single-electron in interface states in the channel of n-type MOSFETs with decanano dimensions using 3D atomistic simulation techniques. In order to highlight the basic dependencies, the simulations are carried out initially assuming continuous doping charge, and discrete localized charge only for the trapped electron. The dependence of the random telegraph signal (RTS) amplitudes on the device dimensions and on the position of the trapped charge in the channel are studied in detail. Later, in full-scale, atomistic simulations assuming discrete charge for both randomly placed dopants and the trapped electron, we highlight the importance of current percolation and of traps with strategic position where the trapped electron blocks a dominant current path. | Automated understanding of spatio-temporal usage patterns of real-world applications are significant in urban planning. With the capability of smartphones collecting various information using inbuilt sensors, the smart city data is enriched with multiple contexts. Whilst tensor factorization has been successfully used to capture latent factors (patterns) exhibited by the real-world datasets, the multifaceted nature of smart city data needs an improved modeling to utilize multiple contexts in sparse condition. Thus, in our ongoing research, we aim to model this data with a novel Context-Aware Nonnegative Coupled Sparse Matrix Tensor (CAN-CSMT) framework which imposes sparsity constraint to learn the true factors in sparse data. We also aim to develop a fast and efficient factorization algorithm to deal with the scalability problem persistent in the state-of-the-art factorization algorithms. | eng_Latn | 886 |
Parallel ALS Algorithm for Solving Linear Systems in the Hierarchical Tucker Representation | Tensor network formats are an efficient tool for numerical computations in many dimensions, yet even this tool often becomes too time- and memory-consuming for a single compute node when applied to problems of scientific interest. Intending to overcome such limitations, we present and analyze a parallelization scheme for algorithms based on the hierarchical Tucker representation which distributes the network vertices and their associated computations over a set of distributed-memory processors. We then propose a modified version of the alternating least squares algorithm for solving linear systems amenable to parallelization according to the aforementioned scheme, and highlight technical considerations important for obtaining an efficient and stable implementation. Our numerical experiments support the theoretical assertion that the parallel scaling of this algorithm is constrained only by the dimensionality and the rank uniformity of the targeted problem. | Many data analysis problems involve an investigation of relationships between attributes in heterogeneous databases, where different prediction models can be more appropriate for different regions. We propose a technique of integrating global and local random subspace ensemble. We performed a comparison with other well known combining methods on standard benchmark datasets and the proposed technique gave better accuracy. | eng_Latn | 887 |
Intrusion Detection Method Based on Improved K-Means Algorithm | Data mining technology has a good application in the field of intrusion detection. For the problem that K-Means algorithm is difficult to process high-dimensional data, local optimal solution, and cannot determine K value, this paper proposes an improved K-Means algorithm. Firstly, the PCA algorithm is used to reduce the dimension of the data set, and then the Outlier detection is used to eliminate the Outliers that have a great influence on the final clustering result. Then, the initial clustering center point is selected based on the distance to avoid the local optimal solution. Finally, K- is used. The Means algorithm performs clustering to obtain an intrusion cluster. The experimental results show that compared with the common data mining-based intrusion detection algorithm, the proposed method has a good performance in detection rate and false positive rate, and its performance has also improved. | We present an incremental approach to 2-norm estimation for triangular matrices. Our investigation covers both dense and sparse matrices which can arise for example from a QR, a Cholesky or a LU factorization. If the explicit inverse of a triangular factor is available, as in the case of an implicit version of the LU factorization, we can relate our results to incremental condition estimation (ICE). Incremental norm estimation (INE) extends directly from the dense to the sparse case without needing the modifications that are necessary for the sparse version of ICE. INE can be applied to complement ICE, since the product of the two estimates gives an estimate for the matrix condition number. Furthermore, when applied to matrix inverses, INE can be used as the basis of a rank-revealing factorization. | eng_Latn | 888 |
Efficient block noise removal based on nonlinear manifolds | The problem of block noise removal is considered. It is assumed that the original image is on or close to a sub-space of admissible images in the form of a low dimensional nonlinear manifold. We propose to use a close variant of the total variation regularizer for measuring block noise. Based on this noise measure, we present an effective approach that reconstructs the original image in the presence of block noise. Our main computational task is the solution of a quadratic programming problem, for which we propose a very efficient interior point method. The effectiveness and efficiency of our approach is demonstrated by an example. | We study the problem of learning a kernel matrix from an apriori kernel and training data. An unconstrained convex optimization formulation is proposed, with an arbitrary convex smooth loss function on kernel entries and a LogDet divergence for regularization. Since the number of variables is of order O(n), standard Newton and quasi-Newton methods are too time-consuming. An operator form Hessian is used to develop an O(n) trust-region inexact Newton method, where the Newton direction is computed using several conjugate gradient steps on the Hessian operator equation. On the uspst dataset, our algorithm can handle 2 million optimization variables within one hour. Experiments are shown for both linear (Mahalanobis) metric learning and for kernel learning. The convergence rate, speed and performance of several loss functions and algorithms are discussed. | eng_Latn | 889 |
Incremental Norm Estimation for Dense and Sparse Matrices | We present an incremental approach to 2-norm estimation for triangular matrices. Our investigation covers both dense and sparse matrices which can arise for example from a QR, a Cholesky or a LU factorization. If the explicit inverse of a triangular factor is available, as in the case of an implicit version of the LU factorization, we can relate our results to incremental condition estimation (ICE). Incremental norm estimation (INE) extends directly from the dense to the sparse case without needing the modifications that are necessary for the sparse version of ICE. INE can be applied to complement ICE, since the product of the two estimates gives an estimate for the matrix condition number. Furthermore, when applied to matrix inverses, INE can be used as the basis of a rank-revealing factorization. | This paper presents a clustering model based on neural network(NN_Cluster) combining the self adaptive feature of neural network,in order to solve the noise data of clustering.Then design two clustering algorithms based on adaptive resonance theory neural network(ARTNN_Cluster) and self-organizing feature map neural networks(SOMNN_Cluster).Simulation results on UCI datasets demonstrate that comparing with traditional K_means clustering means,the NN_Cluster effectively overcome the noise of traditional clustering methods and the better clustering results are obtained by this model. | eng_Latn | 890 |
Tension-Compression Test of a Concrete Specimen Via a Structure Damage Theory | The concrete sample is assumed to consist of a large number of aggregates and cement paste. Their material properties are assumed to be perfectly elastic. The effective elastic property of such a perfectly bonded composite material can be determined. We shall first replace the cement-aggregate composite by a homogeneous elastic material with a set of imaginary interior cement-aggregate boundary surfaces. These surfaces are assumed to be weak in some sense, and the purpose of this paper is to discuss the effects of the presence and propagation of cracks on these surfaces on the overall response of the sample. | Automated understanding of spatio-temporal usage patterns of real-world applications are significant in urban planning. With the capability of smartphones collecting various information using inbuilt sensors, the smart city data is enriched with multiple contexts. Whilst tensor factorization has been successfully used to capture latent factors (patterns) exhibited by the real-world datasets, the multifaceted nature of smart city data needs an improved modeling to utilize multiple contexts in sparse condition. Thus, in our ongoing research, we aim to model this data with a novel Context-Aware Nonnegative Coupled Sparse Matrix Tensor (CAN-CSMT) framework which imposes sparsity constraint to learn the true factors in sparse data. We also aim to develop a fast and efficient factorization algorithm to deal with the scalability problem persistent in the state-of-the-art factorization algorithms. | kor_Hang | 891 |
Integrin signaling: it's where the action is | Recent advances highlight a critical role for integrin receptors for extracellular matrix in determining where in cells critical signals are transduced. Integrins are shown to activate signaling intermediates at specific surface membrane locations, to promote nuclear translocation of factors that activate gene transcription, and to recruit and augment the signaling power of receptors for growth factors. | We present an incremental approach to 2-norm estimation for triangular matrices. Our investigation covers both dense and sparse matrices which can arise for example from a QR, a Cholesky or a LU factorization. If the explicit inverse of a triangular factor is available, as in the case of an implicit version of the LU factorization, we can relate our results to incremental condition estimation (ICE). Incremental norm estimation (INE) extends directly from the dense to the sparse case without needing the modifications that are necessary for the sparse version of ICE. INE can be applied to complement ICE, since the product of the two estimates gives an estimate for the matrix condition number. Furthermore, when applied to matrix inverses, INE can be used as the basis of a rank-revealing factorization. | eng_Latn | 892 |
The nilpotency class of the sandwich subalgebra of simple finite-dimensional Lie algebras | In this paper it is shown that the nilpotency class of sandwich subalgebras in Lie algebras of Cartan type and in Melikyan algebras over a field of characteristic is equal to , where is the sum of the heights of the variables. The only exception is the Zassenhaus algebra (and also the Hamiltonian and contact algebras for ). | Automated understanding of spatio-temporal usage patterns of real-world applications are significant in urban planning. With the capability of smartphones collecting various information using inbuilt sensors, the smart city data is enriched with multiple contexts. Whilst tensor factorization has been successfully used to capture latent factors (patterns) exhibited by the real-world datasets, the multifaceted nature of smart city data needs an improved modeling to utilize multiple contexts in sparse condition. Thus, in our ongoing research, we aim to model this data with a novel Context-Aware Nonnegative Coupled Sparse Matrix Tensor (CAN-CSMT) framework which imposes sparsity constraint to learn the true factors in sparse data. We also aim to develop a fast and efficient factorization algorithm to deal with the scalability problem persistent in the state-of-the-art factorization algorithms. | eng_Latn | 893 |
Method for separating and extracting vanadium and chromium from vanadium chromium leaching liquor | The invention belongs to the field of wet metallurgy, and particularly relates to a method for separating and extracting vanadium and chromium from vanadium chromium leaching liquor. The method comprises the following steps of: adding sodium sulfite into the vanadium chromium leaching liquor, adjusting the pH value to 5 to 5.5; carrying out the precipitation reaction; filtering, calcining filtered solids and cooling; adding the calcination products into NaOH (sodium hydroxide) solution to obtain solid filter residue, i.e. chromium sesquioxide; adjusting the pH value of the liquid filtered after the precipitatio to 2.0 to 2.5, heating the filtered liquid to 90 to 95 DEG C, and adding ammonium sulfate to obtain precipitates, filtering the precipitates, wherein the solid filter residue is ammonium metavanadate. By adopting the moderate reduction, less reduction agent is added, the reduction action is carried out under a normal temperature condition, and the cost is reduced. | Gaussian Process Regression (GPR) and Gaussian Process Latent Variable Models (GPLVM) offer a principled way of performing probabilistic non-linear regression and dimensionality reduction. In this paper we propose a hybrid between the two, the covariate-GPLVM (c-GPLVM), to perform dimensionality reduction in the presence of covariate information (e.g. continuous covariates, class labels, or censored survival times). This construction lets us adjust for covariate effects and reveals meaningful latent structure which is not revealed when using GPLVM. Furthermore, we introduce structured decomposable kernels which will let us interpret how the fixed and latent inputs contribute to feature-level variation, e.g. identify the presence of a non-linear interaction. We demonstrate the utility of this model on applications in disease progression modelling from high-dimensional gene expression data in the presence of additional phenotypes. | eng_Latn | 894 |
What is sparse matrix? | What is a sparse matrix? | What is the difference between rote learning vs meaningful learning? | eng_Latn | 895 |
What is penalized logistic regression | Why does shrinkage work? | The square roots of different primes are linearly independent over the field of rationals | eng_Latn | 896 |
Variational mesh decomposition | A convex relaxation approach for computing minimal partitions | Widespread reward-system activation in obese women in response to pictures of high-calorie foods | eng_Latn | 897 |
A Safe Screening Rule for Sparse Logistic Regression | Additive Logistic Regression: a Statistical View of Boosting | Efficient Euclidean projections in linear time | eng_Latn | 898 |
Minimizing Nonconvex Population Risk from Rough Empirical Risk. | On the Complexity of Bandit and Derivative-Free Stochastic Convex Optimization | Optimization of Discrete Cosine Transform-Based Image Watermarking by Genetics Algorithm | eng_Latn | 899 |