• 대한수학회보
• /
• 제51권2호
• /
• pp.519-530
• /
• 2014

Let (R,m) be a Noetherian local ring and M a finitely generated R-module. For an integer s > -1, we say that M is Cohen-Macaulay in dimension > s if every system of parameters of M is an M-sequence in dimension > s introduced by Brodmann-Nhan [1]. In this paper, we give some characterizations for Cohen-Macaulay modules in dimension > s in terms of the Noetherian dimension of the local cohomology modules $H^i_m(M)$, the polynomial type of M introduced by Cuong [5] and the multiplicity e($\underline{x}$;M) of M with respect to a system of parameters $\underline{x}$.

• Journal of the Korean Statistical Society
• /
• 제29권1호
• /
• pp.9-16
• /
• 2000

This article addresses the problem of maximum likelihood estimation in ARCH regression with lagged dependent variables. Some topics in asymptotics of the model such as uniform expansion of likelihood function and construction of a class of MLE are discussed, and the regularity property of MLE is obtained. The error process here is possibly non-Gaussian.

• Journal of the Korean Statistical Society
• /
• 제18권1호
• /
• pp.46-61
• /
• 1989

This paper is concerned with the development of a nonparametric procedure for the statistical inference about the nonlinear regression parameters. A confidence region and a hypothesis testing procedure based on a class of signed linear rank statistics are proposed and the asymptotic distributions of the test statistic both under the null hypothesis and under a sequence of local alternatives are investigated. Some desirable asymptotic properties including the asymptotic relative efficiency are discussed for various score functions.

• 대한전자공학회논문지SP
• /
• 제38권6호
• /
• pp.665-674
• /
• 2001

본 논문에서는 압축 동 영상의 블록화 및 링 현상을 동시에 제거 할 수 있는 저 계산량의 효율적인 필터방식을 제안한다. 인접 화소와의 상관 관계를 고려한 새로운 1차원 정규화 함수를 정의하여 기존의 정규화 복원 방식이 갖고 있던 계산량의 부하 문제를 해결하였다. 제안된 1차원 정규화 함수는 처리하고자 하는 화 소의 두 인접 화소를 이용한 2개의 부가 함수로 구성되어 있으며, 각각의 정규화 매개 변수는 복호화부에서 이용 가능한 매크로 블록의 타입 및 양자화 스텝 크기 등의 부가 정보를 이용하여 예측한다. 더불어, 본 논문에서는 정규화 매개 변수를 룩업 테이블 (look-up table)로 구성할 수 있도록 정규화 매개 변수의 영역을 제한하여 계산량을 더욱 줄일 수 있도록 구성하였다. 제안된 방식의 효율성을 실험 결과를 통해 확인할 수 있었다.

• 한국생물정보학회:학술대회논문집
• /
• 한국생물정보시스템생물학회 2000년도 International Symposium on Bioinformatics
• /
• pp.28-31
• /
• 2000

We describe a hidden Markov model, HMMTIR, for general protein sequence based on the I-sites library of sequence-structure motifs. Unlike the linear HMMs used to model individual protein families, HMMSTR has a highly branched topology and captures recurrent local features of protein sequences and structures that transcend protein family boundaries. The model extends the I-sites library by describing the adjacencies of different sequence-structure motifs as observed in the database, and achieves a great reduction in parameters by representing overlapping motifs in a much more compact form. The HMM attributes a considerably higher probability to coding sequence than does an equivalent dipeptide model, predicts secondary structure with an accuracy of 74.6% and backbone torsion angles better than any previously reported method, and predicts the structural context of beta strands and turns with an accuracy that should be useful for tertiary structure prediction. HMMSTR has been incorporated into a public, fully-automated protein structure prediction server.

• Interdisciplinary Bio Central
• /
• 제1권2호
• /
• pp.8.1-8.6
• /
• 2009

Introduction: It is a challenge to design a protein score function which stabilizes the native structures of many proteins simultaneously. The coarse-grained description of proteins to construct the pairwise-contact score function usually ignores the backbone directionality of protein structures. We propose a new two-body score function which stabilizes all native states of 1,006 proteins simultaneously. This two-body score function differs from the usual pairwise-contact functions in that it considers two adjacent amino acids at two ends of each peptide bond with the backbone directionality from the N-terminal to the C-terminal. The score is a corresponding propensity for a directional alignment of two adjacent amino acids with their local environments. Results and Discussion: We show that the construction of a directional adjacency-score function was achieved using 1,006 training proteins with the sequence homology less than 30%, which include all representatives of different protein classes. After parameterizing the local environments of amino acids into 9 categories depending on three secondary structures and three kinds of hydrophobicity of amino acids, the 32,400 adjacency-scores of amino acids could be determined by the perceptron learning and the protein threading. These could stabilize simultaneously all native folds of 1,006 training proteins. When these parameters are tested on the new distinct 382 proteins with the sequence homology less than 90%, 371 (97.1%) proteins could recognize their native folds. We also showed using these parameters that the retro sequence of the SH3 domain, the B domain of Staphylococcal protein A, and the B1 domain of Streptococcal protein G could not be stabilized to fold, which agrees with the experimental evidence.

• 자원환경지질
• /
• 제28권4호
• /
• pp.425-431
• /
• 1995

Genetic algorithms are so named because they are analogous to biological processes. The model parameters are coded in binary form. The algorithm then starts with a randomly chosen population of models called chromosomes. The second step is to evaluate the fitness values of these models, measured by a correlation between data and synthetic for a particular model. Then, the three genetic processes of selection, crossover, and mutation are performed upon the model in sequence. Genetic algorithms share the favorable characteristics of random Monte Carlo over local optimization methods in that they do not require linearizing assumptions nor the calculation of partial derivatives, are independent of the misfit criterion, and avoid numerical instabilities associated with matrix inversion. An additional advantage over converntional methods such as iterative least squares is that the sampling is global, rather than local, thereby reducing the tendency to become entrapped in local minima and avoiding the dependency on an assumed starting model.

• Communications for Statistical Applications and Methods
• /
• 제2권2호
• /
• pp.198-212
• /
• 1995

This study provides a nonparametric procedure for the statistical inference of the parameters in stationary autoregressive processes. A confidence region and a hypothesis testing procedure based on a class of signed linear rank statistics are proposed and the asymptotic distributions of the test statistic both underthe null hypothesis and under a sequence of local alternatives are investigated. Some desirable asymptotic properties including the asymptotic relative efficiency are discussed for various score functions.

• Communications for Statistical Applications and Methods
• /
• 제2권2호
• /
• pp.288-295
• /
• 1995

In this paper a hypotheses test procedure based on the least absolute deviation estimators for the unknown parameters in nonlinear regression models is investigated. The asymptotic distribution of the proposed likelihood ratio test statistic are established voth under the null hypotheses and a sequence of local alternative hypotheses. The asymptotic relative efficiency of the proposed test with classical test based on the least squares estimator is also discussed.

• 말소리와 음성과학
• /
• 제8권2호
• /
• pp.23-30
• /
• 2016

This paper describes some implementation schemes of CNN in view of mini-batch DNN training for efficient second order optimization. This uses same procedure updating parameters of DNN to train parameters of CNN by simply arranging an input image as a sequence of local patches, which is actually equivalent with mini-batch DNN training. Through this conversion, second order optimization providing higher performance can be simply conducted to train the parameters of CNN. In both results of image recognition on MNIST DB and syllable automatic speech recognition, our proposed scheme for CNN implementation shows better performance than one based on DNN.