• 제목/요약/키워드: MCMC Method

검색결과 103건 처리시간 0.024초

Bayesian Estimation for Skew Normal Distributions Using Data Augmentation

  • Kim Hea-Jung
    • Communications for Statistical Applications and Methods
    • /
    • 제12권2호
    • /
    • pp.323-333
    • /
    • 2005
  • In this paper, we develop a MCMC method for estimating the skew normal distributions. The method utilizing the data augmentation technique gives a simple way of inferring the distribution where fully parametric frequentist approaches are not available for small to moderate sample cases. Necessary theories involved in the method and computation are provided. Two numerical examples are given to demonstrate the performance of the method.

GPU 를 활용한 스캔라인 블록 Gibbs 샘플링 기법의 가속 (Accelerating Scanline Block Gibbs Sampling Method using GPU)

  • ;김원식;;박인규
    • 한국방송∙미디어공학회:학술대회논문집
    • /
    • 한국방송공학회 2014년도 하계학술대회
    • /
    • pp.77-78
    • /
    • 2014
  • A new MCMC method for optimization is presented in this paper, which is called the scanline block Gibbs sampler. Due to its slow convergence speed, traditional Markov chain Monte Carlo (MCMC) is not widely used. In contrast to the conventional MCMC method, it is more convenient to parallelize the scanline block Gibbs sampler. Since The main part of the scanline block Gibbs sampler is to calculate message between each edge, in order to accelerate the calculation of messages passing in scanline sampler, it is parallelized in GPU. It is proved that the implementation on GPU is faster than on CPU based on the experiments on the OpenGM2 benchmark.

  • PDF

Effects of ILFs on DRAM algorithm in SURR model uncertainty evaluation caused by interpolated rainfall using different methods

  • Nguyen, Thi Duyen;Nguyen, Duc Hai;Bae, Deg-Hyo
    • 한국수자원학회:학술대회논문집
    • /
    • 한국수자원학회 2022년도 학술발표회
    • /
    • pp.137-137
    • /
    • 2022
  • Evaluating interpolated rainfall uncertainty of hydrological models caused by different interpolation methods for basins where can not fully collect rainfall data are necessary. In this study, the adaptive MCMC method under effects of ILFs was used to analyze the interpolated rainfall uncertainty of the SURR model for Gunnam basin, Korea. Three events were used to calibrate and one event was used to validate the posterior distributions of unknown parameters. In this work, the performance of four ILFs on uncertainty of interpolated rainfall was assessed. The indicators of p_factor (percentage of observed streamflow included in the uncertainty interval) and r_factor (the average width of the uncertainty interval) were used to evaluate the uncertainty of the simulated streamflow. The results showed that the uncertainty bounds illustrated the slight differences from various ILFs. The study confirmed the importance of the likelihood function selection in the application the adaptive Bayesian MCMC method to the uncertainty assessment of the SURR model caused by interpolated rainfall.

  • PDF

구분적 선형함수에서의 베이지안 변화점 추출 (Bayesian Detection of Multiple Change Points in a Piecewise Linear Function)

  • 김정연
    • 응용통계연구
    • /
    • 제27권4호
    • /
    • pp.589-603
    • /
    • 2014
  • 본 연구는 시간의 순서에 따라 순차적으로 발생한 신호 자료에 있어서, 변화점 검출을 위한 베이지안 방법을 개발하고자 한다. 특히, Reversible Jump MCMC를 이용하여, 차원이 정해지지 않은 모수 공간을 탐색할 수 있는 효율적인 베이지안 추론 모형을 개발한다. 신호가 각 구간에서 선형함수인 경우에 대한 모형과 이해가 용이한 모형을 제안하고, 추정을 위해 고유의 MCMC알고리즘을 개발하였다. 제안된 방법을 모의실험 자료에 적용함으로써 그 정확성 및 효율성을 검증하였고, 실제 자료에도 적용하여 보았다.

Geostatistics for Bayesian interpretation of geophysical data

  • Oh Seokhoon;Lee Duk Kee;Yang Junmo;Youn Yong-Hoon
    • 한국지구물리탐사학회:학술대회논문집
    • /
    • 한국지구물리탐사학회 2003년도 Proceedings of the international symposium on the fusion technology
    • /
    • pp.340-343
    • /
    • 2003
  • This study presents a practical procedure for the Bayesian inversion of geophysical data by Markov chain Monte Carlo (MCMC) sampling and geostatistics. We have applied geostatistical techniques for the acquisition of prior model information, and then the MCMC method was adopted to infer the characteristics of the marginal distributions of model parameters. For the Bayesian inversion of dipole-dipole array resistivity data, we have used the indicator kriging and simulation techniques to generate cumulative density functions from Schlumberger array resistivity data and well logging data, and obtained prior information by cokriging and simulations from covariogram models. The indicator approach makes it possible to incorporate non-parametric information into the probabilistic density function. We have also adopted the MCMC approach, based on Gibbs sampling, to examine the characteristics of a posteriori probability density function and the marginal distribution of each parameter. This approach provides an effective way to treat Bayesian inversion of geophysical data and reduce the non-uniqueness by incorporating various prior information.

  • PDF

Bayesian Analysis in Generalized Log-Gamma Censored Regression Model

  • Younshik chung;Yoomi Kang
    • Communications for Statistical Applications and Methods
    • /
    • 제5권3호
    • /
    • pp.733-742
    • /
    • 1998
  • For industrial and medical lifetime data, the generalized log-gamma regression model is considered. Then the Bayesian analysis for the generalized log-gamma regression with censored data are explained and following the data augmentation (Tanner and Wang; 1987), the censored data is replaced by simulated data. To overcome the complicated Bayesian computation, Makov Chain Monte Carlo (MCMC) method is employed. Then some modified algorithms are proposed to implement MCMC. Finally, one example is presented.

  • PDF

How to Improve Classical Estimators via Linear Bayes Method?

  • Wang, Lichun
    • Communications for Statistical Applications and Methods
    • /
    • 제22권6호
    • /
    • pp.531-542
    • /
    • 2015
  • In this survey, we use the normal linear model to demonstrate the use of the linear Bayes method. The superiorities of linear Bayes estimator (LBE) over the classical UMVUE and MLE are established in terms of the mean squared error matrix (MSEM) criterion. Compared with the usual Bayes estimator (obtained by the MCMC method) the proposed LBE is simple and easy to use with numerical results presented to illustrate its performance. We also examine the applications of linear Bayes method to some other distributions including two-parameter exponential family, uniform distribution and inverse Gaussian distribution, and finally make some remarks.

그리드 단체 위의 디리슐레 분포에서 마르코프 연쇄 몬테 칼로 표집 (MCMC Algorithm for Dirichlet Distribution over Gridded Simplex)

  • 신봉기
    • 정보과학회 컴퓨팅의 실제 논문지
    • /
    • 제21권1호
    • /
    • pp.94-99
    • /
    • 2015
  • 비모수 베이스 통계학, 확률적 표집에 기반한 추론 등이 기계학습의 주요 패러다임으로 등장하면서 디리슐레(Dirichlet) 분포는 최근 다양한 그래프 모형 곳곳에 등장하고 있다. 디리슐레 분포는 일변수 감마 분포를 벡터 분포로 확장한 형태의 하나이다. 본 논문에서는 감마 분포를 갖는 임의의 자연수 X를 K개의 자연수의 합으로 임의 분할 할 때 각 부분의 크기 비율을 디리슐레 분포에서 표집하는 방법을 제안한다. 일반적으로 디리슐레 분포는 연속적인 (K-1)-단체(simplex) 위에 정의 되지만 자연수로 분할하는 표본은 자연수라는 조건 때문에 단체 내부의 이산 그리드 점에만 정의된다. 본 논문에서는 단체 위의 그리드 상의 이웃 점들의 확률 분포로부터 마르코프연쇄 몬테 칼로(MCMC) 제안 분포를 정의하고 일련의 표본들의 마르코프 연쇄를 구현하는 알고리듬을 제안한다. 본 방법은 마르코프 모델, HMM 및 준-HMM 등에서 각 상태별 시간 지속 분포를 표현하는데 활용 가능하다. 나아가 최근 제안된 전역-지역(global-local) 상태지속 분포를 동시에 모형화하는 감마-디리슐레 HMM에도 응용가능하다.

베이지안 방식에 의한 지구물리 역산 문제의 접근 (A Bayesian Approach to Geophysical Inverse Problems)

  • 오석훈;정승환;권병두;이희순;정호준;이덕기
    • 지구물리와물리탐사
    • /
    • 제5권4호
    • /
    • pp.262-271
    • /
    • 2002
  • 본 연구에서는 지구물리 자료의 베이지안 역산을 효과적으로 수행하는 방법에 관해 논의하였다. 베이지안 처리에서 가장 문제가 되는 사전확률분포를 구하기 위해 지구통계학적 방법을 적용하였으며, 사후확률분포의 추정을 위해 MCMC(Markov Chain Monte Carlo) 방법을 적용하였다. 쌍극자배열 전기비저항 탐사 자료의 2차원 역산을 위해 슐럼버저배열 전기비저항탐사 자료와 시추공 자료를 사전 정보로 이용하였으며, 이들 사전정보에 대해 지구통계학적 방법을 적용하여 사전확률분포를 작성하였다. 쌍극자배열 전기비저항 탐사 자료를 최대 우도함수로 하는 사후확률분포는 차원이 매우 높은 적분을 요구하므로, 이를 추정하기 위해 MCMC기술을 적용하였으며, 보다 효율적인 접근을 위해 Gibbs샘플링 방법을 이용하였다. 그 결과 비모수적 방식으로 사후확률분포를 분석함으로써 보다 신뢰성 있는 해를 구할 수 있었으며, 주변화(marginalization)된 사후확률분포를 이용하여 다양한 분석을 적용할 수 있었다.

Analyze the parameter uncertainty of SURR model using Bayesian Markov Chain Monte Carlo method with informal likelihood functions

  • Duyen, Nguyen Thi;Nguyen, Duc Hai;Bae, Deg-Hyo
    • 한국수자원학회:학술대회논문집
    • /
    • 한국수자원학회 2021년도 학술발표회
    • /
    • pp.127-127
    • /
    • 2021
  • In order to estimate parameter uncertainty of hydrological models, the consideration of the likelihood functions which provide reliable parameters of model is necessary. In this study, the Bayesian Markov Chain Monte Carlo (MCMC) method with informal likelihood functions is used to analyze the uncertainty of parameters of the SURR model for estimating the hourly streamflow of Gunnam station of Imjin basin, Korea. Three events were used to calibrate and one event was used to validate the posterior distributions of parameters. Moreover, the performance of four informal likelihood functions (Nash-Sutcliffe efficiency, Normalized absolute error, Index of agreement, and Chiew-McMahon efficiency) on uncertainty of parameter is assessed. The indicators used to assess the uncertainty of the streamflow simulation were P-factor (percentage of observed streamflow included in the uncertainty interval) and R-factor (the average width of the uncertainty interval). The results showed that the sensitivities of parameters strongly depend on the likelihood functions and vary for different likelihood functions. The uncertainty bounds illustrated the slight differences from various likelihood functions. This study confirms the importance of the likelihood function selection in the application of Bayesian MCMC to the uncertainty assessment of the SURR model.

  • PDF