• 한국수자원학회논문집
• /
• 제42권4호
• /
• pp.331-341
• /
• 2009

본 연구에서는 generalized logistic(GL) 분포의 최우도 추정량(maximum likelihood estimate)에 대한 불확실성 추정을 위하여 사용되는 관측정보행렬(observed information matrix)과 Fisher 정보행렬(Fisher information matrix)의 정확도를 비교해 보고자 하였다. 타 분포형에 대한 기존의 연구결과에서 표본의 크기가 클 경우 매개변수 추정시 관측정보행렬이 동시에 추정되어 계산시간도 단축되고 Fisher 정보행렬의 정확도와도 차이도 거의 없어 관측정보행렬의 사용이 추천된 바 있으나, 최근 사용이 증가되고 있는 GL 분포에 대한 연구결과는 아직 전무한 실정이며 기존 연구문헌의 결과를 토대로 구체적인 연구 없이 관측정보행렬을 사용하고 있는 상황이다. 따라서 본 연구에서는 이를 위해 모의실험을 수행하였으며, 모의 결과 최우도법에 의한 매개변수의 분산 및 공분산은 기존의 연구 결과와 비슷한 결과를 보이나, quantile에 대한 불확실성 추정에는 관측정보행렬보다 Fisher 정보행렬의 사용이 더 적절할 것으로 판단되었다.

• Journal of the Korean Statistical Society
• /
• 제31권2호
• /
• pp.237-250
• /
• 2002

Motivated by Cook's (1986) assessment of local influence by investigating the curvature of a surface associated with the overall discrepancy measure, this paper extends this idea to the linear regression model with AR(p) disturbances. Diagnostic for the linear regression models with AR(p) disturbances are discussed when simultaneous perturbations of the response vector are allowed. For the derived criterion, numerical studies demonstrate routine application of this work.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제11권5호
• /
• pp.2607-2627
• /
• 2017

Nonnegative matrix factorization (NMF) has received considerable attention due to its effectiveness of reducing high dimensional data and importance of producing a parts-based image representation. Most of existing NMF variants attempt to address the assertion that the observed data distribute on a nonlinear low-dimensional manifold. However, recent research results showed that not only the observed data but also the features lie on the low-dimensional manifolds. In addition, a few hard priori label information is available and thus helps to uncover the intrinsic geometrical and discriminative structures of the data space. Motivated by the two aspects above mentioned, we propose a novel algorithm to enhance the effectiveness of image representation, called Dual graph-regularized Constrained Nonnegative Matrix Factorization (DCNMF). The underlying philosophy of the proposed method is that it not only considers the geometric structures of the data manifold and the feature manifold simultaneously, but also mines valuable information from a few known labeled examples. These schemes will improve the performance of image representation and thus enhance the effectiveness of image classification. Extensive experiments on common benchmarks demonstrated that DCNMF has its superiority in image classification compared with state-of-the-art methods.

• ETRI Journal
• /
• 제24권6호
• /
• pp.469-472
• /
• 2002

Speaker adaptation techniques are generally used to reduce speaker differences in speech recognition. In this work, we focus on the features fitted to a linear regression-based speaker adaptation. These are obtained by feature transformation based on independent component analysis (ICA), and the feature transformation matrices are estimated from the training data and adaptation data. Since the adaptation data is not sufficient to reliably estimate the ICA-based feature transformation matrix, it is necessary to adjust the ICA-based feature transformation matrix estimated from a new speaker utterance. To cope with this problem, we propose a smoothing method through a linear interpolation between the speaker-independent (SI) feature transformation matrix and the speaker-dependent (SD) feature transformation matrix. From our experiments, we observed that the proposed method is more effective in the mismatched case. In the mismatched case, the adaptation performance is improved because the smoothed feature transformation matrix makes speaker adaptation using noisy speech more robust.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권6호
• /
• pp.2634-2648
• /
• 2020

In this paper, an adaptive sparse singular value decomposition (ASSVD) algorithm is proposed to estimate the signal matrix when only one data matrix is observed and there is high dimensional white noise, in which we assume that the signal matrix is low-rank and has sparse singular vectors, i.e. it is a simultaneously low-rank and sparse matrix. It is a structured matrix since the non-zero entries are confined on some small blocks. The proposed algorithm estimates the singular values and vectors separable by exploring the structure of singular vectors, in which the recent developments in Random Matrix Theory known as anisotropic Marchenko-Pastur law are used. And then we prove that when the signal is strong in the sense that the signal to noise ratio is above some threshold, our estimator is consistent and outperforms over many state-of-the-art algorithms. Moreover, our estimator is adaptive to the data set and does not require the variance of the noise to be known or estimated. Numerical simulations indicate that ASSVD still works well when the signal matrix is not very sparse.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2009년도 정보 및 제어 심포지움 논문집
• /
• pp.323-325
• /
• 2009

In this paper, we present simulation results of Dynamic Matrix Control(DMC) to a boiler steam temperature. In order to control of steam temperature, we choose the input-output variables and generate the step response model by each input variable's step test. After that, the control structure executes on-line control with optimization using step response model. Proposed controller is applied to the APESS(Doosan company's boiler model simulator) and it is observed that the simulation results show satisfactory performance of proposed control.

• Journal of the Korean Data and Information Science Society
• /
• 제21권3호
• /
• pp.547-554
• /
• 2010

In this paper, we extend the work by Lee et al. (2010) to multidimensional di usion processes. A test statistic analogous to the one-dimensional case is proposed to inves-tigate the joint stability of covariance matrix parameters and, under certain regularity conditions, is shown to have a limiting distribution of the sup of a multidimensional Brownian bridge. A simulation result is provided for illustration.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권7호
• /
• pp.3778-3793
• /
• 2019

Image steganography is one of the key types of steganography where a message to be sent is hidden inside the cover image. The most commonly used techniques for image steganography rely on LSB steganography. In this paper, a novel image steganography technique based on blocks matrix determinant method is proposed. Under this method, a cover image is divided into blocks of size $2{\times}2$ pixels and the determinant of each block is calculated. The comparison of the determinant values and corresponding data bits yields a delicate way for the embedment of data bits. The main aim of the proposed technique is to ensure concealment of secret data inside an image without affecting the cover image quality. When the proposed steganography method is compared with other existing LSB steganography methods, it is observed that it not only provides higher PSNR, lower MSE but also guarantees better quality of the stego image.

• ETRI Journal
• /
• 제42권6호
• /
• pp.922-931
• /
• 2020

A hierarchical iterative algorithm for the canonical polyadic decomposition (CPD) of tensors is proposed by improving the traditional conjugate gradient least squares (CGLS) method. Methods based on algebraic operations are investigated with the objective of estimating the direction of arrival (DoA) and polarization parameters of signals impinging on an array with electromagnetic (EM) vector-sensors. The proposed algorithm adopts a hierarchical iterative strategy, which enables the algorithm to obtain a fast recovery for the highly collinear factor matrix. Moreover, considering the same accuracy threshold, the proposed algorithm can achieve faster convergence compared with the alternating least squares (ALS) algorithm wherein the highly collinear factor matrix is absent. The results reveal that the proposed algorithm can achieve better performance under the condition of fewer snapshots, compared with the ALS-based algorithm and the algorithm based on generalized eigenvalue decomposition (GEVD). Furthermore, with regard to an array with a small number of sensors, the observed advantage in estimating the DoA and polarization parameters of the signal is notable.

• 반도체디스플레이기술학회지
• /
• 제16권4호
• /
• pp.36-40
• /
• 2017

With an attempt to enhance the visibility of laser beam, we have investigated a black matrix with scattering particles by ray tracing simulations. As the scattering particle density is increased, the detected power by the receiver is increased, thereby enhancing the visibility. In reality, the visibility is reduced with increasing incident angle (away from the normal incidence) of laser beam, a phenomenon also observed by ray tracing simulations. It is due to the fact that the mean path is increased within a highly absorptive BM layer or a smaller number of rays hit the BM area when the incident angle is high. Embedding a number of scattering particles into BM may bring in crosstalk among pixels. However, it is negligible because scattered rays inside highly absorptive BM are re-scattered due to the high scattering particle density, decreasing the power of scattered rays into the active areas.