• ETRI Journal
• /
• v.39 no.1
• /
• pp.21-29
• /
• 2017

This paper addresses the problem of unsupervised speech separation based on robust non-negative matrix factorization (RNMF) with ${\beta}$-divergence, when neither speech nor noise training data is available beforehand. We propose a robust version of non-negative matrix factorization, inspired by the recently developed sparse and low-rank decomposition, in which the data matrix is decomposed into the sum of a low-rank matrix and a sparse matrix. Efficient multiplicative update rules to minimize the ${\beta}$-divergence-based cost function are derived. A convolutional extension of the proposed algorithm is also proposed, which considers the time dependency of the non-negative noise bases. Experimental speech separation results show that the proposed convolutional RNMF successfully separates the repeating time-varying spectral structures from the magnitude spectrum of the mixture, and does so without any prior training.

• Journal of the Korean Data and Information Science Society
• /
• v.28 no.4
• /
• pp.927-936
• /
• 2017

To analyze longitudinal count data, Poisson linear mixed models are commonly used. In the models the random effects covariance matrix explains both within-subject variation and serial correlation of repeated count outcomes. When the random effects covariance matrix is assumed to be misspecified, the estimates of covariates effects can be biased. Therefore, we propose reasonable and flexible structures of the covariance matrix using autoregressive and moving average Cholesky decomposition (ARMACD). The ARMACD factors the covariance matrix into generalized autoregressive parameters (GARPs), generalized moving average parameters (GMAPs) and innovation variances (IVs). Positive IVs guarantee the positive-definiteness of the covariance matrix. In this paper, we use the ARMACD to model the random effects covariance matrix in Poisson loglinear mixed models. We analyze epileptic seizure data using our proposed model.

• Journal of the Korean Statistical Society
• /
• v.25 no.1
• /
• pp.49-66
• /
• 1996

The singular value decomposition is one of the most useful methods in the area of matrix computation. It gives dimension reduction which is the centeral idea in many multivariate analyses. But this method is not resistant, i.e., it is very sensitive to small changes in the input data. In this article, we derive the resistant version of singular value decomposition for principal component analysis. And we give its statistical applications to biplot which is similar to principal component analysis in aspects of the dimension reduction of an n x p data matrix. Therefore, we derive the resistant principal component analysis and biplot based on the resistant singular value decomposition. They provide graphical multivariate data analyses relatively little influenced by outlying observations.

• Journal of Institute of Control, Robotics and Systems
• /
• v.6 no.6
• /
• pp.462-470
• /
• 2000

This paper proposes a visual servoing method based on QR decomposition and disturbance observer. The QR decomposition factors the image feature Jacobian into a unitary matrix and an upper triangular matrix. And it is shown that several performance indices such as measurement sensitivity of visual features, sensitivity of the control to noise and controllability can be improved for any general image feature Jacobian by QR decomposition and disturbance observer. To show the validity of the proposed approach, visual servoing with stereo vision is carried out for a Samsung FARAMAN 6-axis industrial robot manipulator.

• Journal of applied mathematics & informatics
• /
• v.6 no.2
• /
• pp.523-538
• /
• 1999

A new preconditioning method is investigated to reduce the roundoff error in computing derivatives using Chebyshev col-location methods(CCM). Using this preconditioning causes ration of roundoff error of preconditioning method and CCm becomes small when N gets large. Also for accuracy enhancement of differentiation we use a domain decomposition approach. Error analysis shows that for this domain decomposition method error reduces proportional to the length of subintervals. Numerical results show that using domain decomposition and preconditioning simultaneously gives super accu-rate approximate values for first derivative of the function and good approximate values for moderately high derivatives.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.21 no.2
• /
• pp.103-109
• /
• 2021

About the orthogonal Hadamard matrix announced by Hadamard in France in 1893, Professor Moon Ho Lee newly defined it as Center Weight Hadamard in 1989 and announced it, and discovered the Jacket matrix in 1998. The Jacket matrix is a generalization of the Hadamard matrix. In this paper, we propose a method of obtaining the Symmetric Jacket matrix, analyzing important properties and patterns, and obtaining the Jacket matrix's determinant and Eigenvalue, and proved it using Eigen decomposition. These calculations are useful for signal processing and orthogonal code design. To analyze the matrix system, compare it with DFT, DCT, Hadamard, and Jacket matrix. In the symmetric matrix of Galois Field, the element-wise inverse relationship of the Jacket matrix was mathematically proved and the orthogonal property AB=I relationship was derived.

• Journal of computational fluids engineering
• /
• v.3 no.1
• /
• pp.22-29
• /
• 1998

Parallel implementation is conducted for a SIMPLER finite volume model. The present parallelism is based on domain decomposition and explicit message passing using MPI and SHMEM. Two parallel solvers to tridiagonal matrix equation are employed. The implementation is verified on the Cray T3E system for a benchmark problem of natural convection in a sidewall-heated cavity. The test results illustrate good scalability of the present parallel models. Performance issues are elaborated in view of convergence as well as conventional parallel overheads and single processor performance. The effectiveness of a localized matrix solution algorithm is demonstrated.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2013.10a
• /
• pp.130-132
• /
• 2013

This paper presents an improvement for calculating method of astronomical vessel position with circle of equal altitude equation based on using a virtual object in sun and two stars observation. In addition, to enhance the accuracy of ship position achieved from solving linear matrix system, and surmount the disadvantages on rank deficient matrices situation, the authors used singular value decomposition (SVD) in least square method instead of normal equation and QR decomposition, so, the solution of matrix system will be available in all situation. As proposal algorithm, astronomical ship position will give more accuracy than previous methods.

• Journal of Ship and Ocean Technology
• /
• v.7 no.3
• /
• pp.56-65
• /
• 2003

The hull monitoring systems of container ships with four long-base gages give enough information for identifying the hull girder loads such as bending and torsional moments. But such a load-identification for container ships has not been known. In this paper, a load-identification method is suggested in terms of a linear matrix equation that the measured strain vector equals to the multiplication of the transformation matrix and the desired strain component vector. The equation is proved to be mathematically complete by the property of positive-definite determinant of the transformation matrix. The method is applied to a hull stress monitoring system for 8100TED container ship during sea trial, and the estimated external loads illustrate reasonable results in comparison with the pre-estimated results. This moment decomposition concept has also been tested in real operation conditions. The typical phenomena over the Suez Canal illustrated very suitable results comparing with the physical understandings. Henceforth, one can effectively use the proposed concept to monitor the hull girder loads such as bending and torsional moments.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• v.1
• /
• pp.133-138
• /
• 2006

Singular value decomposition (SDV) approach is applied to the observability analysis of GPS/INS in this paper. A measure of observability for a subspace is introduced. It indicates the minimum size of perturbation in the information matrix that makes the subspace unobservable. It is shown that the measure has direct connections with observability of systems, error covariance, and singular structure of the information matrix. The observability measure given in this paper is applicable to the multi-input/multi-output time-varying systems. An example on the observability analysis of GPS/INS is given. The measure of observability is confirmed to be less sensitive to system model perturbation. It is also shown that the estimation error for the vertical component of gyro bias can be considered unobservable for small initial error covariance for a constant velocity horizontal motion.