• Journal of Korea Multimedia Society
• /
• v.11 no.9
• /
• pp.1302-1309
• /
• 2008

Traditional clustering methods, like k-means or fuzzy clustering, are prototype-based methods which are applicable only to convex clusters. On the other hand, spectral clustering tries to find clusters only using local similarity information. Its ability to handle concave clusters has gained the popularity recent years together with support vector machine (SVM) which is a kernel-based classification method. However, as is in SVM, the kernel width plays an important role and has a great impact on the result. Several methods are proposed to decide it automatically, it is still determined based on heuristics. In this paper, we proposed an adaptive method deciding the kernel width based on distance histogram. The proposed method is motivated by the fact that the affinity matrix should be formed into a block diagonal matrix to generate the best result. We use the tradition Euclidean distance together with the random walk distance, which make it possible to form a more apparent block diagonal affinity matrix. Experimental results show that the proposed method generates more clear block structured affinity matrix than the existing one does.

• The Korean Journal of Applied Statistics
• /
• v.35 no.4
• /
• pp.569-577
• /
• 2022

Simulation studies are often conducted when it is difficult to compare the performance of nonparametric estimators theoretically. Kim and Kim (2021) showed that more systematic and accurate comparisons can be made if you analyze the simulation results using a regression model,. This study is a complementary study on Kim and Kim (2021). In the variance-covariance matrix for the error term of the regression model, only heteroscedasticity was considered and covariance was ignored in the previous study. When covariance is considered together with the heteroscedasticity, the variance-covariance matrix becomes a block diagonal matrix. In this study, a method of estimating and using the block diagonal variance-covariance matrix for the analysis was presented. This allows you to find more pairs of estimators with significant performance differences while ensuring the nominal confidence level.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.35 no.11A
• /
• pp.1027-1037
• /
• 2010

In this paper, we propose the linear interpolation-based BD (Block Diagonalization) precoding approximation algorithm for low complexity in downlink multiuser MIMO-OFDM (Multiple-input Multiple-output Orthogonal Frequency Division Multiplexing) systems. In the case of applying the general BD precoding algorithm to multiuser MIMO-OFDM systems, the computational complexity increases in proportional to the number of subcarriers. The proposed interpolation-based BD precoding approximation algorithm can be achieved similar SER performance with general BD algorithm and can decrease the computational complexity. It is proved that proposed algorithm can achieve the significantly decreased computational complexity by computer simulation.

• Proceedings of the Korea Multimedia Society Conference
• /
• 2003.11b
• /
• pp.994-997
• /
• 2003

행렬 연산은 계산 과학을 사용하는 공학 물리, 화학, 생명 과학, 경제학 등에서 다양하게 사용되고 있으며 이 행렬은 크기가 크고 대부분의 원소가 0 값을 갖는 희소 행렬일 경우가 많다. 본 논문에서는 희소 행렬의 연산 중, 회로 설계 시 최적화 과정에 사용되는 연산에서 문제가 되는 희소 행렬 A 와 블록 대각 행렬 H에 대하여 AH$A^{T}$ 의 연산을 효율적으로 행하는 방법들을 검토하고 메모리 접근 횟수를 모델링하여 수행 속도와 메모리 사용량 면에서 비교한다.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.33 no.9
• /
• pp.34-40
• /
• 2005

In present study, a two dimensional unsteady Navier-Stokes solver has been developed using the Diagonalized ADI (DADI) method and implicit dual time stepping method. The jacobian matrices in steady state Navier-Stokes equations are introduced from inviscid flux terms. The implicit treatment of artificial dissipation terms results in a block penta-diagonal matrix system and it becomes a scalar penta-diagonal matrix by diagonalization. In steady state equations about fictitious time, a new residual including a real time derivative term is introduced. From a converged solution about fictitious time, a real time unsteady solution can be obtained, which is called 'implicit dual time stepping method'. For code validation, an oscillating flat plate, a regular Karman vortices past a circular cylinder and shock buffeting around a bicircular airfoil problems are numerically solved. And they are compared with a theoretical solution, experiments and other researcher's computations.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.41 no.3
• /
• pp.231-238
• /
• 2004

This paper presents the high speed noise reduction processing system using the modified discrete fourier transform(MDFT) on the frequency domain. The proposed filter uses the linear prediction coefficients of the adaptive line enhance(ALE) method based on the Sign algorithm The signals with a random noise tracking performance are examined through computer simulations. It is confirmed that the fast adaptive digital filter is realized by the high speed adaptive noise reduction(HANR) algorithm with rapid convergence on the frequency domain(FD).

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.6
• /
• pp.634-639
• /
• 2020

In general, the stability and response characteristics of the system can be improved by changing the pole position because a nonlinear system can be linearized by the product of a 1st and 2nd order system. Therefore, a controller that moves the pole can be designed in various ways. Among the other methods, LQ control ensures the stability of the system. On the other hand, it is difficult to specify the location of the pole arbitrarily because the desired response characteristic is obtained by selecting the weighting matrix by trial and error. This paper evaluated a method of selecting a weighting matrix of LQ control that moves multiple double poles with Jordan blocks to real poles. The relational equation between the double poles and weighting matrices were derived from the characteristic equation of the Hamiltonian system with a diagonal control weighting matrix and a state weighting matrix represented by two variables (ρ_d, ϕ_d). The Moving-Range was obtained under the condition that the state-weighting matrix becomes a positive semi-definite matrix. This paper proposes a method of selecting poles in this range and calculating the weighting matrices by the relational equation. Numerical examples are presented to show the usefulness of the proposed method.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.1
• /
• pp.20-27
• /
• 2020

In general, a nonlinear system is linearized in the form of a multiplication of the 1st and 2nd order system. This paper reports a design method of a weighting matrix and control law of LQ control to move the double poles that have a Jordan block to a pair of complex conjugate poles. This method has the advantages of pole placement and the guarantee of stability, but this method cannot position the poles correctly, and the matrix is chosen using a trial and error method. Therefore, a relation function (𝜌, 𝜃) between the poles and the matrix was derived under the condition that the poles are the roots of the characteristic equation of the Hamiltonian system. In addition, the Pole's Moving-range was obtained under the condition that the state weighting matrix becomes a positive semi-definite matrix. This paper presents examples of how the matrix and control law is calculated.

• Journal of the Korea Society of Computer and Information
• /
• v.11 no.3
• /
• pp.141-149
• /
• 2006

In this paper, we introduce an improved algorithm for computing matrix triple product that commonly arises in primal-dual optimization method. In computing $P=AHA^{t}$, we devise a single pass algorithm that exploits the block diagonal structure of the matrix H. This one-phase scheme requires fewer floating point operations and roughly half the memory of the generic two-phase algorithm, where the product is computed in two steps, computing first $Q=HA^{t}$ and then P=AQ. The one-phase scheme achieved speed-up of 2.04 on Intel Itanium II platform over the two-phase scheme. Based on memory latency and modeled cache miss rates, the performance improvement was evaluated through performance modeling. Our research has impact on performance tuning study of complex sparse matrix operations, while most of the previous work focused on performance tuning of basic operations.

• The Journal of the Acoustical Society of Korea
• /
• v.30 no.6
• /
• pp.315-323
• /
• 2011

This paper proposes two methods of estimating prior distribution to improve the performance of rapid speaker adaptation based on maximum a posteriori linear regression (MAPLR). In general, prior distribution of the transformation matrix used in MAPLR adaptation is estimated from all of the training speakers who are employed to construct the speaker-independent model, and it is applied identically to all new speakers. In this paper, we propose a method in which prior distribution is estimated from a group of reference speakers, selected using adaptation data, so that the acoustic characteristics of the selected reference speakers may be similar to that of the new speaker. Additionally, in MAPLR adaptation with block-diagonal transformation matrix, we propose a method in which the mean matrix and covariance matrix of prior distribution are estimated from two groups of transformation matrices obtained from the same training speakers, respectively. To evaluate the performance of the proposed methods, we examine word accuracy according to the number of adaptation words in the isolated word recognition task. Experimental results show that, for very limited adaptation data, statistically significant performance improvement is obtained in comparison with the conventional MAPLR adaptation.