• The Korean Journal of Applied Statistics
• /
• v.11 no.2
• /
• pp.435-449
• /
• 1998

In this article we have introduced a Bayes criterion for the variable selection in multiple discriminant analysis (MDA). The criterion is a default Bayes factor for the comparision of homo/heteroscadasticity of the multivariate normal means. The default Bayes factor is obtained from a development of the imaginary training sample method introduced by Spiegelhalter and Smith (1982). Based an the criterion, we also provided a test for additional discrimination in MDA. The advantage of the criterion is that it is not only applicable for the optimal subset selection method but for the stepwise method. More over, the criterion can be reduced to that for two-group discriminant analysis. Thus the criterion can be regarded as an unified alternative to variable selection criteria suggested by various sampling theory approaches. To illustrate the performance of the criterion, a numerical study has bean done via Monte Carlo experiment.

• Proceedings of the Korean Nuclear Society Conference
• /
• 1998.05a
• /
• pp.273-278
• /
• 1998

Belief network(or Bayesian network) based on Bayes' rule in probabilistic theory can be applied to the reasoning of diagnostic systems. This paper describes the basic theory of concept and feasibility of using the network for diagnosis of nuclear power plants. An example shows that the probabilities of root causes of a failure are calculated from the measured or believed evidences.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1993.06a
• /
• pp.806-809
• /
• 1993

This paper focuses on the usage of the fuzzy set theory in decision making systems. The approach to calculation of generalized membership function, based on application of method of principal components is proposed. For solving of the problem of fuzzy forecasting the development of Bayes procedure is used. The structure of decision making system, in which following procedures are fulfilled, is discussed.

• Annual Conference on Human and Language Technology
• /
• 2001.10d
• /
• pp.465-471
• /
• 2001

본 연구에서는 동형이의어 분별을 위하여 허정(2000)이 제시한 "사전 뜻풀이말에서 추출한 의미정보에 기반한 동형이의어 중의성 해결 시스템" 이 가지는 문제점과 향후 연구과제로 제시한 문제들을 개선하기 위하여 Bayes 정리에 기반한 동형이의어 분별 모델을 제안한다. 의미 분별된 사전 뜻풀이말 코퍼스에서 동형이의어를 포함하고 있는 뜻풀이말을 구성하는 체언류(보통명사), 용언류(형용사, 동사) 및 부사류(부사)를 의미 정보로 추출한다. 동형이의어의 의미별 사전 출현 빈도수가 비교적 균등한 기존 9개의 동형이의어 명사를 대상으로 실험하여 비교하였고, 새로 7개의 동형이의어 용언(형용사, 동사)을 추가하여 실험하였다. 9개의 동형이의어 명사를 대상으로 한 내부 실험에서 평균 99.37% 정확률을 보였으며 1개의 동형이의어 용언을 대상으로 한 내부 실험에서 평균 99.53% 정확률을 보였다. 외부 실험은 국어 정보베이스와 ETRI 코퍼스를 이용하여 9개의 동형이의어 명사를 대상으로 평균 84.42% 정확률과 세종계획의 350만 어절 규모의 외부 코퍼스를 이용하여 7개의 동형이의어 용언을 대상으로 평균 70.81%의 정확률을 보였다.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.49 no.1
• /
• pp.53-58
• /
• 2012

With a game-theoretic view, p-persistence slotted ALOHA is structured as a non-cooperative game, in which a Nash equilibrium is sought to provide a value for the probability of attempting to deliver a packet. An expression of Nash equilibrium necessarily includes the number of active outer stations, which is hardly available in many practical applications. In this paper, we thus propose a Bayesian scheme of predicting the number of active outer stations prior to deciding whether to attempt to deliver a packet or not. Despite only requiring the minimal information that an outer station is genetically able to acquire by itself, the Bayesian scheme demonstrates the competitive predicting performance against a method which depends on heavy information.

• Journal of Information Science Theory and Practice
• /
• v.2 no.3
• /
• pp.29-39
• /
• 2014

In recent years, several studies have proposed making use of the Twitter micro-blogging service to track various trends in online media and discussion. In this study, we specifically examine the use of Twitter to track discussions of food safety in the Korean language. Given the irregularity of keyword use in most tweets, we focus on optimistic machine-learning and feature set selection to classify collected tweets. We build the classifier model using Naive Bayes & Naive Bayes Multinomial, Support Vector Machine, and Decision Tree Algorithms, all of which show good performance. To select an optimum feature set, we construct a basic feature set as a standard for performance comparison, so that further test feature sets can be evaluated. Experiments show that precision and F-measure performance are best when using a Naive Bayes Multinomial classifier model with a test feature set defined by extracting Substantive, Predicate, Modifier, and Interjection parts of speech.

• International Journal of Reliability and Applications
• /
• v.13 no.1
• /
• pp.37-47
• /
• 2012

In this article we attempted reliability analysis of a component under the stress-strength pattern with both classical as well as Bayesian techniques. The main focus is made to develop the theory for dealing the reliability problems in various circumstances for bivariate environmental set up in context of Bayesian paradigm. A stress-strength based model describes the life of a component which has strength (Y) and is subjected to stress(X). We develop the Bayes and moment estimators of reliability of a component for each of the three possible conditions, under the assumption that the two stresses (i.e. $X_1$ and $X_2$) on a component are dependent and follow a Bivariate exponential (BVE) of Marshall-Olkin distribution, the strength of a component (Y) following exponential distribution is independent of the stresses. The simulation study is performed with Markov Chain Monte Carlo technique via Gibbs sampler to obtain the estimates of Bayes estimators of reliability, are compared with moment estimators of reliabilities on the basis of absolute biases.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2005.11a
• /
• pp.117-120
• /
• 2005

퍼지 규칙 기반 시스템에서 분류 및 경계를 결정하기 위한 방법으로 퍼지 규칙을 학습하는 다양한 방법들이 제안되고 있다. 그리고 추론 규칙간의 상관성을 고려하여 불필요한 속성을 제거함으로써 좀 더 효율적인 추론 결과를 얻을 수 있다. 따라서 본 논문에서는 퍼지 규칙 기반 시스템에서 각 규칙에 따른 결정 테이블를 작성하고 러프집합을 이용하여 불필요한 속성을 제거하였으며 규칙의 확신도에 퍼지 네이브 베이스 이론을 적용한 추론 방법을 제안한다.

• IEIE Transactions on Smart Processing and Computing
• /
• v.6 no.1
• /
• pp.18-20
• /
• 2017

In this paper, we propose a new classification method to complement a $na{\ddot{i}}ve$ Bayesian classifier. This classifier assumes data distribution to be Gaussian, finds the discriminant function, and derives the decision curve. However, this method does not investigate finding the decision curve in much detail, and there are some minor problems that arise in finding an accurate discriminant function. Our findings also show that this method could produce errors when finding the decision curve. The aim of this study has therefore been to investigate existing problems and suggest a more reliable classification method. To do this, we utilize the gradient to find the decision curve. We then compare/analyze our algorithm with the $na{\ddot{i}}ve$ Bayesian method. Performance evaluation indicates that the average accuracy of our classification method is about 10% higher than $na{\ddot{i}}ve$ Bayes.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2014.11a
• /
• pp.1136-1138
• /
• 2014

Recently, predicting personality with the help of smartphone usage is become very interesting and attention grabbing topic in the field of research. At present there are some approaches towards detecting a user's personality which uses the smartphones usage data, such as call detail records (CDRs), the usage of short message services (SMSs) and the usage of social networking services application. In this paper, we focus on the predicting user attitude based on MBTI theory by using their smartphone usage data. We used Naïve Bayes and SVM classifier for classifying user personalities by extracting some features from smartphone usage data. From analysis it is observed that, SVM classifier works well as compared to Naïve Bayes.