• Journal of the Korean Statistical Society
• /
• 제8권1호
• /
• pp.15-21
• /
• 1979

A zone of sensitivity is developed to investigate the effect of nonnormality on the Bayes decision function for testing mean of a normal population when either parent or prior belongs to Edgeworthian family of moderately nonnormal probability density functions.

• Communications for Statistical Applications and Methods
• /
• 제4권3호
• /
• pp.787-793
• /
• 1997

It is well known that the assumption of independence is ofter not valid for real data. This phenomenon has been observed empirically by many prominent scientists. In this article the sensitivity of dependence on Bayes test of a sharp null hypothesis is considered. The robustness is considered with respect to the significant level and the prior probability on the null hypothesis.

• 한국건설관리학회논문집
• /
• 제17권5호
• /
• pp.13-21
• /
• 2016

주수급자 역할을 하는 건설기업의 부실화는 발주자에게 공사계약 미이행에 따른 피해를 초래할 수 있고, 전문건설업체 및 자재공급업체의 재무건전성에 악영향을 줄 수 있다. 건설업은 프로젝트를 수주하고 진도에 따라 기성을 받는 현금흐름의 재무적 특성이 존재하고, 사업 진행 중의 부실화는 투입한 자금의 손실로 이어질 수 있으므로 건설업체의 부실화 예측은 중요하다. 국내 건설업체의 부실화 예측은 90년도 초 미국에서 개발된 KMV (Kealhofer McQuown and Vasicek)사의 KMV모형으로 수행되는 경우도 있지만, 이 모형은 일반적인 기업 및 은행의 신용위험 평가에 개발되어져 건설기업 예측력에는 부족함이 있다. 또한, KMV값의 부도확률 예측력에 대해서는 분석대상의 기업수 및 데이터의 부족으로 의문점이 지속적으로 제기되고 있다. 따라서 이러한 의문점을 해결하기 위해 기존 부도예측확률모형에 베이지안 확률적 접근법(Bayesian Probabilistic Approach)을 접목하고자 한다. 베이즈 통계학의 사전확률(Prior Probability)만 적절하게 예측가능하다면 적은 정보라도 증거에 대한 조건부 획득으로 신뢰성 있는 사후확률(Posterior Probability)을 예측할 수 있기 때문이다. 이에 본 연구에서는 기존 부도예측확률모형에 베이지안 확률적 접근법을 활용하여 예상부도확률(Expected Default Frequency, EDF)을 측정하고, 기존 모형의 예상부도확률과 비교하여 정확성을 예측하고자 한다.

• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2015년도 춘계학술대회
• /
• pp.593-594
• /
• 2015

In wireless sensor networks (WSNs), due to the very low data rate, the sleeping schedule is usually used to save consumed energy and prolong the lifetime of nodes. However, duty-cycled approach can cause a high end-to-end (E2E) delay. In this paper, we study a node scheduling algorithm in WSNs such that E2E delay meets bounded delay with a given probability. We have applied the probability theory to spot the relationship between E2E delay and node interval. Simulation result illustrates that we can create the network to achieve given delay with prior probability and high energy use efficient as well.

• ETRI Journal
• /
• 제27권4호
• /
• pp.461-464
• /
• 2005

Recently, a variety of confidence measures for utterance verification has been studied to improve speech recognition performance by rejecting out-of-vocabulary inputs. Most of the conventional confidence measures for utterance verification are based primarily on hypothesis testing or an approximated posterior probability, and their performances depend on the robustness of an alternative hypothesis or the prior probability. We introduce a novel confidence measure called a search confusion rate (SCR), which does not require an alternative hypothesis or the approximation of posterior probability. Our confusion-based approach shows better performance in additive noise-corrupted speech as well as in clean speech.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제10권10호
• /
• pp.5095-5111
• /
• 2016

The Probability Hypothesis Density (PHD) filter is a suboptimal approximation and tractable alternative to the multi-target Bayesian filter based on random finite sets. However, the PHD filter fails to track newborn targets when the target birth intensity is unknown prior to tracking. In this paper, a dual detection-guided newborn target intensity PHD algorithm is developed to solve the problem, where two schemes, namely, a newborn target intensity estimation scheme and improved measurement-driven scheme, are proposed. First, the newborn target intensity estimation scheme, consisting of the Dirichlet distribution with the negative exponent parameter and target velocity feature, is used to recursively estimate the target birth intensity. Then, an improved measurement-driven scheme is introduced to reduce the errors of the estimated number of targets and computational load. Simulation results demonstrate that the proposed algorithm can achieve good performance in terms of target states, target number and computational load when the newborn target intensity is not predefined in multi-target tracking systems.

• 대한전자공학회논문지SP
• /
• 제41권5호
• /
• pp.111-116
• /
• 2004

웨이블릿 영역에서 Bayesian 추정법을 이용한 잡음 제거를 위해서는 웨이블릿 계수의 prior 모델, 잡음의 확률분포, 웨이블릿 계수에 대한 분산 등의 정보가 필요하다. 잡음 제거의 일반적인 방법은 웨이블릿 계수에 대한 적절한 prior 모델을 설정하고 이에 대한 신호의 분산을 추정하는 것이다. 본 논문에서는 영역 확장 방법을 사용하여 영상의 영역에 따라 분산을 추정하기 위한 창의 크기를 결정하는 방법을 제안한다. 이웃 계수의 범위는 동질성 척도를 정의하여 가장 작은 영역부터 영역을 확장하는 방법을 사용한다. 결정된 가변 이웃 영역을 사용하여 원 신호의 분산을 결정하고 이를 이용하여 웨이블릿 영역에서 Bayesian 추정법을 사용하여 영상의 잡음을 제거한다. 실험 결과를 통하여 제안 방법이 기존의 방법보다 높은 PSNR을 나타냄을 보여 준다.

• 한국임상약학회지
• /
• 제25권1호
• /
• pp.9-17
• /
• 2015

Background: Biologic disease-modifying antirheumatic drugs (bDMARDs) extend the treatment choices for rheumatoid arthritis patients with insufficient response or intolerance to conventional DMARDs (cDMARDs). These agents have considerable efficacy compared with conventional DMARDs, but only a few head-to-head comparisons among these agents have been performed. The objective of this systematic review and network meta-analysis (NMA) was to compare the relative efficacy of Certolizumab with conventional DMARD to licensed bDMARD with cDMARD therapy for patients who failed to prior cDMARD treatment under the condition of the reimbursement coverage criteria in Korea. Methods: A systematic review was conducted using MEDLINE and Cochrane library. Key endpoints were the American College of Rheumatology (ACR) responses of 20/50/70 at six months. Bayesian outcomes were calculated as median of treatment effect, probability of the best, Odds Ratio (OR) and probability that OR was greater than one. Results: Compared with other bDMARDs, Certolizumab were associated with higher or comparable ACR response rates; in ACR20, the OR (probability of OR>1) was 2.08 (92.6%) for Adalimumab, 1.86 (85.7%) for Etanercept, 1.89 (79.5%) for Golimumab, 2.36 (92.1%) for Infliximab, 1.79 (87.0%) for Abatacept, 1.74 (80.8%) for Rituximab and 1.82 (86.8%) for Tocilizaumab. In ACR50 and ACR70, the ORs did not present significant differences. Conclusion: Certolizaumab with cDMARD was more effective or comparable than other bDMARDs in patients who failed prior cDMARD treatment.

• 한국사회복지학
• /
• 제62권4호
• /
• pp.249-274
• /
• 2010

본 연구는 한국복지패널 1~3차년도 자료를 이용하여 우울과 국민기초생활보장제도 수급지위 변화의 관계를 분석하였다. 분석 결과는 다음과 같다. 첫째, 수급진입은 우울수준을 증가시킨다. 이는 주로 빈곤진입을 동반한 수급진입이 경제적 스트레스를 증가시켰기 때문으로 판단된다. 둘째, 수급탈출은 우울수준을 감소시키지 않는다. 하지만 수급탈출 중에서도 탈빈곤을 동반한 수급탈출의 경우에는 우울수준을 감소시킨다. 셋째, 높은 우울수준이 수급진입확률을 증가시킨다. 넷째, 높은 우울수준이 수급탈출 확률을 감소시키지 않는다. 수급탈출을 나누어 보면, 높은 우울수준은 탈출 후에도 여전히 빈곤한 수급탈출(탈락)의 가능성을 증가시킨다.

• 한국농공학회논문집
• /
• 제49권3호
• /
• pp.11-20
• /
• 2007

Estimation of crop consumptive use is a key term of agricultural water resource systems design and operation. The 10-year return period drought has special aspects as a reference period in design process of irrigation systems in terms of agricultural water demand analysis so that crop evapotranspiration (ETc) about the return period also has to be analyzed to assist understanding of crop water requirement of paddy rice. In this study, The ETc of 10-year return period drought was computed using frequency analysis by 54 meteorological stations. To find an optimal probability distribution, 8 types of probability distribution function were tested by three the goodness of fit tests including ${\chi}^2$(Chi-Square), K-S (Kolmogorov-Smirnov) and PPCC (Probability Plot Correlation Coefficient). Optimal probability distribution function was selected the 2-parameter Log-Normal (LN2) distribution function among 8 distribution functions. Using the two selected distribution functions, the ETc of 10-year return period drought was estimated for 54 meteorological stations and compared with prior study results suggested by other researchers.