• 정보처리학회논문지B
• /
• 제11B권1호
• /
• pp.117-124
• /
• 2004

본 논문은 NaiveBayes에서 정확도의 손실 없이 효율적으로 동작하는 NaiveBayes에 대한 새로운 알고리즘을 제안한다. 제안된 방법은 분류 벡터에 대한 행렬 전치를 사용하여 NaiveBayes의 확률 계산 량을 최소화하는 것이다. 제안된 방법을 문서 분류 프레임 인 AI::Categorizer 상에서 구현하였으며, 잘 알려진 로이터-21578 데이터를 사용하여 기존의 NaiveBayes 방법과 비교하였다. 성능 비교의 결과로부터 제안된 방법이 기존의 NaiveBayes 방법보다 실행 속도측면에서 약 2배 정도의 성능 개선 효과가 있음을 알 수 있었다. 수 있었다.

• 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.

• Communications for Statistical Applications and Methods
• /
• 제7권2호
• /
• pp.549-562
• /
• 2000

Default Bayes factors are computed to test the equality of one Poisson population mean and the equality of two independent Possion population means. As default priors are assumed Jeffreys priors, noninformative improper priors, and default Bayes factors such as three intrinsic Bayes factors of Berger and Pericchi(1996, 1998), the arithmetic, the median, and the geometric intrinsic Bayes factor, and the factional Bayes factor of O'Hagan(1995) are computed. The testing results by each default Bayes factor are compared with those by the classical method in the simulation study.

• 한국컴퓨터정보학회논문지
• /
• 제23권6호
• /
• pp.1-7
• /
• 2018

A multi-label classification is to find multiple labels associated with the input pattern. Multi-label classification can be achieved by extending conventional single-label classification. Common extension techniques are known as Binary relevance, Label powerset, and Classifier chains. However, most of the extended multi-label naive bayes classifier has not been able to accurately estimate posterior probabilities because it does not reflect the label dependency. And the remaining extended multi-label naive bayes classifier has a problem that it is unstable to estimate posterior probability according to the label selection order. To estimate posterior probability well, we propose a new posterior probability estimation method that reflects the probability between all labels and labels efficiently. The proposed method reflects the correlation between labels. And we have confirmed through experiments that the extended multi-label naive bayes classifier using the proposed method has higher accuracy then the existing multi-label naive bayes classifiers.

• Journal of the Korean Statistical Society
• /
• 제27권1호
• /
• pp.11-23
• /
• 1998

A Bayes criterion for testing the equality of covariance matrices of two multivariate normal distributions is proposed and studied. Development of the criterion invloves calculation of Bayes factor using the imaginary sample method introduced by Spiegelhalter and Smith (1982). The criterion is designed to develop a Bayesian test criterion, so that it provides an alternative test criterion to those based upon asymptotic sampling theory (such as Box's M test criterion). For the constructed criterion, numerical studies demonstrate routine application and give comparisons with the traditional test criteria.

• Communications for Statistical Applications and Methods
• /
• 제3권3호
• /
• pp.215-223
• /
• 1996

By using Tukey's generalized lambda distribution, appoximate posterior density is derived under the Bayes-empirical Bayes model. The sensitivity of posterior distribution to the hyperprior distribution is examined by using Tukey's generalized lambda distriburion which approximate many well-knmown distributions. Based upon Monte Varlo simulation studies it can be said that posterior distribution is sensitive to the cariance of the prior distribution and to the symmetry of the hyperprior distribution. Also posterior distribution is approximately obtained by using the following methods : Lindley method, Laplace method and Gibbs sampler method.

• Journal of the Korean Data and Information Science Society
• /
• 제15권1호
• /
• pp.173-181
• /
• 2004

In this paper, we study and examine the sensitivity of the fuzzy-Bayes method whose properties are relatively not known much. Two fuzzy conditions and two actions are considered. Also some normal distributions and uniform distributions are assumed as a prior distribution for a parameter in the fuzzy-Bayes method.

• Journal of information and communication convergence engineering
• /
• 제17권4호
• /
• pp.239-245
• /
• 2019

Recently, effort to obtain various information from the vast amount of social network services (SNS) big data generated in daily life has expanded. SNS big data comprise sentences classified as unstructured data, which complicates data processing. As the amount of processing increases, a rapid processing technique is required to extract valuable information from SNS big data. We herein propose a system that can extract human sentiment information from vast amounts of SNS unstructured big data using the naïve Bayes algorithm and natural language processing (NLP). Furthermore, we analyze the effectiveness of the proposed method through various experiments. Based on sentiment accuracy analysis, experimental results showed that the machine learning method using the naïve Bayes algorithm afforded a 63.5% accuracy, which was lower than that yielded by the NLP method. However, based on data processing speed analysis, the machine learning method by the naïve Bayes algorithm demonstrated a processing performance that was approximately 5.4 times higher than that by the NLP method.

• 한국수자원학회논문집
• /
• 제50권7호
• /
• pp.489-502
• /
• 2017

본 연구에서는 충주댐 유역에 대해 다목적 댐 예측유입량 산정기법 BAYES-ESP를 개발하고 평가하였다. BAYES-ESP 기법은 기존 ESP (Ensemble Streamflow Prediction) 기법에 베이지안 이론을 적용하여 개발하였으며, 수문모델은 ABCD를 활용하였다. 입력자료는 기온, 강수량 자료와 댐 관측유입량 자료를 활용하였으며, 기온 및 강수량은 기상청, 국토교통부, 한국수자원공사의 지점관측자료, 댐 관측유입량은 한국수자원공사의 자료를 이용하였다. 적용성 평가방법은 시계열 분석과 Skill Score를 활용하였으며, 평가기간은 1986~2015년이다. 시계열 분석 결과 ESP 댐 예측유입량(ESP)는 매년 전망값의 큰 차이가 없었으며, 다우년 및 과우년의 예측성이 떨어지는 것으로 나타났다. BAYES-ESP 댐 예측유입량(BAYES-ESP)는 ESP가 관측유입량에 비해 과소모의하는 경향을 보정하였으며, 특히 다우년에 개선효과가 있는 것으로 나타났다. 월별 평균 댐 관측유입량과의 Skill Score 비교분석결과 ESP는 1~3월에 SS가 비교적 높은 값을 보였으며, 나머지 월에는 음의 값을 나타내었다. BAYES-ESP는 ESP와 관측 값 간의 선형적 관계를 갖는 1~3월에 ESP의 정확도를 향상시키는 것으로 나타났다. ESP 기법은 국내 강수특성상 우리나라에 적용하기에는 한계가 있었으며, 이를 개선한 BAYES-ESP 기법은 댐 유입량 예측연구에 가치가 있다고 판단된다.

• Communications for Statistical Applications and Methods
• /
• 제10권2호
• /
• pp.479-496
• /
• 2003

We investigate logit confidence intervals for the odds ratio based on the delta method. These intervals are constructed using pseudo-Bayes estimators. The Gart method and Agresti method smooth the observed counts toward the model of equiprobability and independence, respectively. We obtain better coverage probability by smoothing the observed counts toward the pseudo-Bayes estimators in 2$\times$2 table. We also improve legit confidence intervals in 2$\times$2$\times$K tables by generalizing these ideas. Utilizing pseudo-Bayes estimators, we obtain better coverage probability by smoothing the observed counts toward the conditional independence model, no three-factor interaction model and saturated model in 2$\times$2$\times$K tables.