• 한국산학기술학회논문지
• /
• 제13권2호
• /
• pp.817-821
• /
• 2012

본 논문에서는 환경음 인식에 GMM(Gaussain mixture model)을 이용할 때 MDL(minimum description length)와 BIC(Bayesian information criterion) 모델선택 기준을 이용하여 최적의 혼합모델 개수를 결정하는 방법에 대해 다루었다. 실험은 모두 9가지 종류의 환경음으로부터 12차 MFCC(mel-frequency cepstral coefficients) 특징 27747개를 추출하고 이를 GMM으로 분류하였다. 각 환경음 클래스의 최적 혼합모델 개수를 추정 하기위해 MDL과 BIC를 적용하고 그 결과를 고정 개수의 혼합모델을 사용한 경우와 비교하였다. 실험 결과에 따르면 혼합모델 선택 방법을 적용한 경우가 그렇지 않은 경우에 비해 거의 유사한 인식성능을 유지하면서 계산복잡도는 BIC와 MDL를 통해 각각 17.8%와 31.7%가 감소하는 것을 확인하였다. 이는 GMM을 이용한 환경음 인식에서 BIC와 MDL 적용을 통해 계산복잡도를 효과적으로 감소시킬 수 있음을 보여준다.

• 한국인터넷방송통신학회논문지
• /
• 제10권6호
• /
• pp.145-152
• /
• 2010

본 연구에서는 시계열 자료를 베이지안 정보기준을 통해 클러스터링 한다. 보다 안정적인 클러스터를 생산하기 위해 다중 중심점을 모델링한 후 이를 이용하여 클러스터를 생성시킨다. 대상 시계열 자료에 대해 예측할 경우 클러스터에 속한 시계열 자료 중 가장 유사한 시계열 자료를 선택하여 모델링한다. 모델로부터 마코프 규칙을 유도해 내고 이 규칙을 이용해 예측정확도를 측정한다. 시계열 자료를 단독으로 모델링한 후 예측한 결과보다 클러스터에 속한 유사시계열 모델링을 통한 예측정확도가 좀 더 높았음을 확인하였다.

• Communications for Statistical Applications and Methods
• /
• 제27권3호
• /
• pp.301-311
• /
• 2020

Schwarz's Bayesian information criterion (BIC) is one of the most popular criteria for model selection, that was derived under the assumption of independent and identical distribution. For correlated data in longitudinal studies, Jones (Statistics in Medicine, 30, 3050-3056, 2011) modified the BIC to select the best linear mixed effects model based on the effective sample size where the number of parameters in covariance structure was not considered. In this paper, we propose an extended Jones' modified BIC by considering covariance parameters. We conducted simulation studies under a variety of parameter configurations for linear mixed effects models. Our simulation study indicates that our proposed BIC performs better in model selection than Schwarz's BIC and Jones' modified BIC do in most scenarios. We also illustrate an example of smoking data using a longitudinal cohort of cancer patients.

• Journal of the Korean Data and Information Science Society
• /
• 제10권1호
• /
• pp.135-146
• /
• 1999

In this paper, we consider the Bayesian hypotheses testing for independence and symmetry in Freund's bivariate exponential model. In Bayesian testing problem, we use the noninformative priors for parameters which are improper and are defined only up to arbitrary constants. And we use the recently proposed hypotheses testing criterion called the intrinsic Bayes factor. Also we derive the arithmetic and median intrinsic Bayes factors and use these results to analyze some data sets.

• Journal of the Korean Data and Information Science Society
• /
• 제17권2호
• /
• pp.291-300
• /
• 2006

In this research, the performance of widely used Bayesian estimators such as Bayes estimator, empirical Bayes estimator, constrained Bayes estimator and constrained empirical Bayes estimator are compared by means of a measurement under balanced loss function for the typical normal-normal situation. The proposed measurement is a weighted sum of the precisions of first and second moments. As a result, one can gets the criterion according to the size of prior variance against the population variance.

• Journal of the Korean Data and Information Science Society
• /
• 제16권2호
• /
• pp.371-382
• /
• 2005

The paper compares the performance of some widely used Bayesian estimators such as Bayes estimator, empirical Bayes estimator, constrained Bayes estimator and constrained Bayes estimator by means of a new measurement under squared error loss function for the typical normal-normal situation. The proposed measurement is a weighted sum of the precisions of first and second moments. As a result, one can gets the criterion according to the size of prior variance against the population variance.

• Communications for Statistical Applications and Methods
• /
• 제12권1호
• /
• pp.139-147
• /
• 2005

It is interesting to locate homogeneous segments within a DNA sequence. Suppose that the DNA sequence has segments within which the observations follow the same residue frequency distribution, and between which observations have different distributions. In this setting, change points correspond to the end points of these segments. This article explores the use of a binary segmentation procedure in detecting the change points in the DNA sequence. The change points are determined using a sequence of nested hypothesis tests of whether a change point exists. At each test, we compare no change-point model with a single change-point model by using the Bayesian information criterion. Thus, the method circumvents the computational complexity one would normally face in problems with an unknown number of change points. We illustrate the procedure by analyzing the genome of the bacteriophage lambda.

• Communications for Statistical Applications and Methods
• /
• 제29권1호
• /
• pp.65-83
• /
• 2022

Although various statistical methods have been developed to map time-dependent genetic factors, most identified genetic variants can explain only a small portion of the estimated genetic variation in longitudinal traits. Gene-gene and gene-time/environment interactions are known to be important putative sources of the missing heritability. However, mapping epistatic gene-gene interactions is extremely difficult due to the very large parameter spaces for models containing such interactions. In this paper, we develop a Gaussian process (GP) based nonparametric Bayesian variable selection method for longitudinal data. It maps multiple genetic markers without restricting to pairwise interactions. Rather than modeling each main and interaction term explicitly, the GP model measures the importance of each marker, regardless of whether it is mostly due to a main effect or some interaction effect(s), via an unspecified function. To improve the flexibility of the GP model, we propose a novel grid-based method for the within-subject dependence structure. The proposed method can accurately approximate complex covariance structures. The dimension of the covariance matrix depends only on the number of fixed grid points although each subject may have different numbers of measurements at different time points. The deviance information criterion (DIC) and the Bayesian predictive information criterion (BPIC) are proposed for selecting an optimal number of grid points. To efficiently draw posterior samples, we combine a hybrid Monte Carlo method with a partially collapsed Gibbs (PCG) sampler. We apply the proposed GP model to a mouse dataset on age-related body weight.

• Journal of the Korean Data and Information Science Society
• /
• 제17권2호
• /
• pp.559-568
• /
• 2006

In this paper, we develop the noninformative priors for the linear mixed models when the parameter of interest is the ratio of variance components. We developed the first and second order matching priors. We reveal that the one-at-a-time reference prior satisfies the second order matching criterion. It turns out that the two group reference prior satisfies a first order matching criterion, but Jeffreys' prior is not first order matching prior. Some simulation study is performed.

• 한국신뢰성학회지:신뢰성응용연구
• /
• 제14권4호
• /
• pp.220-224
• /
• 2014

A Bayesian zero-failure reliability demonstration test method for products with Weibull lifetime distribution is presented. Inverted gamma prior distribution for the scale parameter of the Weibull distribution is used to design the Bayesian test plan and selecting a prior distribution using a prior test information is discussed. A test procedure with zero-failure acceptance criterion is developed that guarantee specified reliability of a product with given confidence level. An example is provided to illustrate the use of the developed Bayesian reliability demonstration test method.