• Communications for Statistical Applications and Methods
• /
• 제4권3호
• /
• pp.817-825
• /
• 1997

Images reconstructed with EM algorithm have been observed to have checkerboard effects and have large distortions near edges as iterations proceed. We suggest a aimple algorithm of applying line process to the EM and Bayesian EM to reduce the distortions near edges. We show by simulation that this algorithm improves the clarity of the reconstructed image and has good properties based on root mean square error.

• Communications for Statistical Applications and Methods
• /
• 제8권2호
• /
• pp.417-426
• /
• 2001

This paper addresses the nonparametric Bayesian estimation for the exponential populations under type II censoring. The Dirichlet process prior is used to provide nonparametric Bayesian estimates of parameters of exponential populations. In the past, there have been computational difficulties with nonparametric Bayesian problems. This paper solves these difficulties by a Gibbs sampler algorithm. This procedure is applied to a real example and is compared with a classical estimator.

• 방송공학회논문지
• /
• 제18권1호
• /
• pp.31-42
• /
• 2013

본 논문에서는 고품질의 3D 콘텐츠 제작에 있어 입체피로를 최소화하기 위한 영상의 수직시차 교정방법으로, 베이지안 접근방식을 적용한 사진측량기반의 강인 편위수정 기법을 제안하고자 한다. 영상의 수직시차 제거 과정은 크게 기하추정 단계와 에피폴라 변환 단계로 구성된다. 본 논문에서는 기하추정을 위해 사진측량에서 널리 활용되고 있는 공면조건 기반의 상대표정 알고리즘을 적용한다. 이때 상대표정 알고리즘에는 자동 정합점 추출에 따른 오정합과 위치오차에 강인성을 확보하기 위해 제약조건을 도입한 베이지안 접근방식을 적용하고자 하며, 이를 바탕으로 수행되는 에피폴라 변환에는 영상의 왜곡과 원 영상 대비 변형을 최소화하기 위한 공선조건기반의 중심투영변환기법을 적용하고자 한다. 알고리즘의 성능검증을 위한 비교 알고리즘으로, 기하추정에는 일반적인 상대표정 알고리즘과 컴퓨터비전분야의 8점 알고리즘 및 스테레오 캘리브레이션 기법이 사용되었으며, 에피폴라 변환에는 Hartley 방법과 Bouguet 방법이 사용되었다. 실험결과는 제안 알고리즘의 높은 정확도와 여러 오차요인들에 대한 강인성, 그리고 최소화된 영상변형의 결과를 보여주었다.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제7권4호
• /
• pp.285-294
• /
• 2007

In this paper, we investigate a novel online estimation algorithm for dynamic Bayesian network(DBN) parameters, given as conditional probabilities. We sequentially update the parameter adjustment rule based on observation data. We apply our algorithm to two well known representations of DBNs: to a first-order Markov Chain(MC) model and to a Hidden Markov Model(HMM). A sliding window allows efficient adaptive computation in real time. We also examine the stochastic convergence and stability of the learning algorithm.

• Smart Structures and Systems
• /
• 제21권4호
• /
• pp.435-448
• /
• 2018

This paper reports the development of a theoretically rigorous method for permanent way engineers to assess the condition of railway ballast under a concrete sleeper with the potential to be extended to a smart system for long-term health monitoring of railway ballast. Owing to the uncertainties induced by the problems of modeling error and measurement noise, the Bayesian approach was followed in the development. After the selection of the most plausible model class for describing the damage status of the rail-sleeper-ballast system, Bayesian model updating is adopted to calculate the posterior PDF of the ballast stiffness at various regions under the sleeper. An obvious drop in ballast stiffness at a region under the sleeper is an evidence of ballast damage. In model updating, the model that can minimize the discrepancy between the measured and model-predicted modal parameters can be considered as the most probable model for calculating the posterior PDF under the Bayesian framework. To address the problems of non-uniqueness and local minima in the model updating process, a two-stage hybrid optimization method was developed. The modified evolutionary algorithm was developed in the first stage to identify the important regions in the parameter space and resulting in a set of initial trials for deterministic optimization to locate all most probable models in the second stage. The proposed methodology was numerically and experimentally verified. Using the identified model, a series of comprehensive numerical case studies was carried out to investigate the effects of data quantity and quality on the results of ballast damage detection. Difficulties to be overcome before the proposed method can be extended to a long-term ballast monitoring system are discussed in the conclusion.

• 약학회지
• /
• 제41권2호
• /
• pp.195-202
• /
• 1997

The purpose of this study was to determine pharmacokinetic parameters of vancomycin using two point calculation(TPC) and Bayesian methods in 16 Korean normal volunteers and 15 g astric cancer patients. Nonparametric expected maximum(NPEM) algorithm for calculation of population pharmacokinetic parameter was used, and these parameters were applied for clinical pharmacokinetic parameters by Bayesian analysis. Vancomycin was administered 1.0g every 12 hrs for 3 days by IV infusion over 60 minutes. The volume of distribution(Vd), elimination rate constant(Kel) and total body clearance(CLt) of vancomycin in normal volunteers using TPC method were $0.34{\pm}0.06 L/kg,\; 0.19{\pm}0.01 hr^{-1}$ and $4.08 {\pm} 0.93 L/hr$, respectively, The Vd, Kel and CLt of vancomycin in gastric cancer patients using TPC method were $0.46 {\pm} 0.06 L/kg, 0.17{\pm}0.02 hr^{-1}$ and $4.84 {\pm} 0.57 L/hr$ respectively. There were significant differences(p<0.05) in Vd. Kel and CLt between normal volunteers and gastric cancer patients. Polpulation pharmacokinetic parameter, the slope(KS) of the relationship beetween Kel versus creatinine Clearance, and the Vd were $0.00157{\pm}0.00029(hr{\cdot}mL/min/1.73m^2)^{-1},\; 0.631 {\pm} 0.0036 L/kg$ in gastric cancer patients using NPEM algorithm respectively. The Vd and Kel were $0.63{\pm}0.005 L/kg, 0.15 {\pm}0.027 hr^{-1}$ for gastric cancer patients using Bayesian method. There were significant differences(p<0.05) in vancomycin pharmacokinetics between Bayesian and TPC methods. It is considered that the population parameter in the patient population is necessary for effective Bayesian method in clinical pharmacy practise.

• 약학회지
• /
• 제43권2호
• /
• pp.160-168
• /
• 1999

The purpose of this study was to determine pharmacokinetic parameters of vancomycin using peak and trough plasma level (PTL) and Bayesian analysis in 20 Korean normal volunteers, 16 gastric cancer and 12 lymphoma patients and also using the compartment model dependent (nonlinear least squares regression: NLSR) and compartment model independent (Lagrange) analysis in 10 ovarian cancer patients. Nonparametric expected maximum (NPEM) algorithm for calculation of the population pharmacokinetic parameters was used, and these parameters were applied for clinical pharmacokinetic parameters by Bayesian analysis. Vancomycin was administered as dose of 1.0 g every 12 hrs for 3 days by IV infusion over 60 minutes in normal volunteers, gastric cancer and lymphoma patients. Population pharmacokinetic parameters, K and Vd in gastric cancer and lymphoma patients using NPEM algorithm were $0.158{\pm}0.014{\;}hr^{-1},{\;}0.630{\pm}0.043{\;}L/kg{\;}and{\;}0.131{\pm}0.0261{\;}hr^{-1},{\;}0.631{\pm}0.089{\;}L/kg$ respectively. The K and Vd in gastric cancer and lymphoma patients using Bayesian analysis were $0.151{\pm}0.027,{\;}0.126{\pm}0.056{\;}hr^{-1}{\;}and{\;}0.62{\pm}0.105,{\;}0.63{\pm}0.095{\;}L/kg$. The K and Vd in ovarian cancer patient using the NLSR and Lagrange analysis were $0.109{\pm}0.008,{\;}0.126{\pm}0.012{\;}hr^{-1}{\;}and{\;} 0.76{\pm}0.08,{\;}0.69{\pm}0.19{\;}L/kg$, respectively. It is necessary for effective dosage regimen of vancomycin in cancer patients to use these population parameters.

• 한국정보전자통신기술학회논문지
• /
• 제8권5호
• /
• pp.400-404
• /
• 2015

본 논문에서는 공간에서 원하는 신호를 추정하기 위해서 도래방향 MUSIC 공간 스펙트럼 알고리즘에 대해서 연구한다. 본 연구에서 제안하는 MUSIC 공간 스펙트럼 알고리즘은 모델 오차와 베이즈 정리를 적용한 방법으로 목표물의 위치를 정확히 추정한다. 적응 배열 안테나를 사용한 수신기의 배열 응답 벡터는 베이즈 방법을 이용하고 모델 오차방법으로 수신 신호의 가중치를 갱신하여 원하는 목표물의 도래 방향을 정확히 추정한다. 본 연구에서 원하는 도래방향 목표물의 신호 추정은 입사 신호의 간섭과 잡음을 제거한 후 배열 응답 벡터를 신호 공분산 행렬의 가중치에 적용한다. 모의실험을 통해서 본 논문에서 제안한 방법과 기존의 도래방향 알고리즘을 비교 분석한다.

• 대한의용생체공학회:의공학회지
• /
• 제22권4호
• /
• pp.311-319
• /
• 2001

Statistical reconstruction methods in the context of a Bayesian framework have played an important role in emission tomography since they allow to incorporate a priori information into the reconstruction algorithm. Given the ill-posed nature of tomographic inversion and the poor quality of projection data, the Bayesian approach uses regularizers to stabilize solutions by incorporating suitable prior models. In this work we show that, while the quantitative performance of the standard filtered backprojection (FBP) algorithm is not as good as that of Bayesian methods, the application of spline-regularized smoothing to the sinogram space can make the FBP algorithm improve its performance by inheriting the advantages of using the spline priors in Bayesian methods. We first show how to implement the spline-regularized smoothing filter by deriving mathematical relationship between the regularization and the lowpass filtering. We then compare quantitative performance of our new FBP algorithms using the quantitation of bias/variance and the total squared error (TSE) measured over noise trials. Our numerical results show that the second-order spline filter applied to FBP yields the best results in terms of TSE among the three different spline orders considered in our experiments.

• 한국정보과학회논문지:소프트웨어및응용
• /
• 제31권12호
• /
• pp.1569-1580
• /
• 2004

베이지안 네트워크는 불확실한 상황을 모델링하기 위한 확률 기반의 모델로서 확실한 수학적 토대를 가지고 있다. 베이지안 네트워크의 구조론 자동 학습하기 위한 연구가 많이 있었고, 최근에는 진화 알고리즘을 이용한 연구가 많이 진행되고 있다. 그러나 대부분은 마지막 세대의 가장 좋은 개체만을 이용하고 있다. 시스템이 요구하는 다양한 요구 조건을 하나의 적합도 평가 수식으로 나타내기 어렵기 때문에, 마지막 세대의 가장 좋은 개체는 종종 편향되거나 변화하는 환경에 덜 적응적일 수 있다. 본 논문에서는 적합도 공유 방법으로 다양한 베이지안 네트워크를 생성하고, 이를 베이즈 규칙을 통해 결합하여 변화하는 환경에 적응적인 추론 모델을 구축할 수 있는 방법을 제안한다. 성능 평가를 위해 ASIA와 ALARM 네트워크에서 인공적으로 생성한 데이타를 이용한 구조 학습 및 추론 실험을 수행하였다. 다양한 조건에서 학습된 네트워크를 실험한 결과, 제안한 방법이 변화하는 환경에서 더욱 강건하고 적응적인 모델을 생성할 수 있음을 알 수 있었다.