• Journal of the Korean Statistical Society
• /
• 제26권4호
• /
• pp.477-491
• /
• 1997

A new predictive classification rule for assigning future cases into one of two multivariate normal population (with unknown normal mixture model) is considered. The development involves calculation of posterior probability of each possible normal-mixture model via a default Bayesian test criterion, called intrinsic Bayes factor, and suggests predictive distribution for future cases to be classified that accounts for model uncertainty by weighting the effect of each model by its posterior probabiliy. In this paper, our interest is focused on constructing the classification rule that takes care of uncertainty about the types of covariance matrices (homogeneity/heterogeneity) involved in the model. For the constructed rule, a Monte Carlo simulation study demonstrates routine application and notes benefits over traditional predictive calssification rule by Geisser (1982).

• Journal of Communications and Networks
• /
• 제17권1호
• /
• pp.75-83
• /
• 2015

In mobile ad-hoc networks (MANETs), nodes are mobile in nature. Collaboration between mobile nodes is more significant in MANETs, which have as their greatest challenges vulnerabilities to various security attacks and an inability to operate securely while preserving its resources and performing secure routing among nodes. Therefore, it is essential to develop an effective secure routing protocol to protect the nodes from anonymous behaviors. Currently, game theory is a tool that analyzes, formulates and solves selfishness issues. It is seldom applied to detect malicious behavior in networks. It deals, instead, with the strategic and rational behavior of each node. In our study,we used the dynamic Bayesian signaling game to analyze the strategy profile for regular and malicious nodes. This game also revealed the best actions of individual strategies for each node. Perfect Bayesian equilibrium (PBE) provides a prominent solution for signaling games to solve incomplete information by combining strategies and payoff of players that constitute equilibrium. Using PBE strategies of nodes are private information of regular and malicious nodes. Regular nodes should be cooperative during routing and update their payoff, while malicious nodes take sophisticated risks by evaluating their risk of being identified to decide when to decline. This approach minimizes the utility of malicious nodes and it motivates better cooperation between nodes by using the reputation system. Regular nodes monitor continuously to evaluate their neighbors using belief updating systems of the Bayes rule.

• 컴퓨터교육학회논문지
• /
• 제21권3호
• /
• pp.51-58
• /
• 2018

본 연구에서는 컴퓨터 오락추구 행동의 예측 분석을 목적으로 한국아동 청소년패널조사(KCYPS) 데이터를 대상으로 베이지안 추론을 사용하여 컴퓨터 오락추구 행동과 관련 변수들의 상호의존성과 인과관계를 조사하였다. 이를 위해 일반 베이지안 네트워크를 통한 마코프 블랭킷(Markov Blanket)을 추출하였다. 또한 변수들의 확률을 변화시켜 컴퓨터 오락추구 행동에 대한 변수들의 영향 정도를 분석하였다. 연구결과, 컴퓨터 오락추구 행동은 관련 변수들(학교학습활동, 비행-흡연, 비행-조롱, 팬덤활동, 학교규칙)의 값을 조정하였을 때 유의미하게 변화되는 것으로 나타났다. 본 연구의 결과로 베이지안 추론은 청소년의 컴퓨터 오락추구 행동을 예측하고 조절하는 등 교육 분야에서 활용될 수 있음을 제시하였다.

• Journal of the Korean Statistical Society
• /
• 제11권1호
• /
• pp.25-35
• /
• 1982

A problem of selecting the least probable cell in a multinomial distribution is studied in a Bayesian framework. We consider two loss components the cost of sampling and the difference in cell probabilities between the selected and the least probable cells. A Bayes sequential selection rule is derived with respect to a Dirichlet prior, and it is compared with the best fixed sample size selection rule. The continuation sets with respect to the vague prior are tabulated for certain cases.

• 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.

• 한국지능정보시스템학회:학술대회논문집
• /
• 한국지능정보시스템학회 2001년도 The Pacific Aisan Confrence On Intelligent Systems 2001
• /
• pp.336-342
• /
• 2001

One of the most important problems on rule induction methods is that they cannot extract rules, which plausibly represent experts decision processes. On one hand, rule induction methods induce probabilistic rules, the description length of which is too short, compared with the experts rules. On the other hand, construction of Bayesian networks generates too lengthy rules. In this paper, the characteristics of experts rules are closely examined and a new approach to extract plausible rules is introduced, which consists of the following three procedures. First, the characterization of decision attributes (given classes) is extracted from databases and the classes are classified into several groups with respect to the characterization. Then, two kinds of sub-rules, characterization rules for each group and discrimination rules for each class in the group are induced. Finally, those two parts are integrated into one rule for each decision attribute. The proposed method was evaluated on a medical database, the experimental results of which show that induced rules correctly represent experts decision processes.

• 한국지구과학회지
• /
• 제24권5호
• /
• pp.428-439
• /
• 2003

여러 지질재해 중에서 산사태로부터 피해를 최소화하기 위해서는 미래의 산사태에 대해 취약한 지역의 추정이 필요하다. 산사태 위험성의 정량적 분석을 목적으로, 본 논문에서는 확률론적 공간통합 방법인 베이지안 기법의 적용가능성에 대해서 논의하고자 한다. 우선 산사태 발생과 관련이 있는 여러 공간자료의 확률론적 표현을 위해 우도비 함수를 사용하였으며, 베이지안 결합 규칙을 이용하여 최종적으로 통합된 검증을 수행하였다. 이러한 방법의 적용가능성을 검토하기 위하여 1998년 여름 산사태 공간 분포의 분할을 통한 검증을 수행하였다. 이러한 방법의 적용가능성을 검토하기 위하여 1998년 여름 산사태로 피해를 입은 경기도 장흥지역을 대상으로 사례연구를 수행하였다. 사례연구 수행 결과, 우도비에 기반한 베이지안 공간 통합 기법은 효율적으로 다양한 공간 자료를 통합할 수 있었으며, 검증결과는 해석과 의사결정 보조자료로 이용될 수 있을 것으로 기대된다.

• 한국정보처리학회논문지
• /
• 제6권7호
• /
• pp.1841-1848
• /
• 1999

델보는 주어진 사례의 집합으로부터 이들 사례들을 분류할 수 있는 베이지안 분류 규칙들로 이루어진 규칙 집합을 습득하는 유전학 기반 귀납적 학습 시스템이다. 규칙 집합의 습득과정에서 델보가 당면하게 되는 한 가지 문제점은 학습 과정이 최적의 규칙 집합이 아닌 지역 최적치를 습득하고 종료하는 경우가 가끔 발생한다는 것이다. 다른 하나의 문제점은 훈련 사례에 대한 경우와는 달리 새로운 평가 사례에 대해 분류 성능이 현저히 저하되는 규칙 집합을 습득하는 경우가 가끔 발생한다는 것이다. 본 논문에서는 이러한 문제점을 해결하여 보다 성능이 향상된 분류 시스템을 구축하기 위한 기법으로 엔-버전 시스템을 구축함으로써 분류 시스템의 전체적인 성능을 향상시키는 기법이다. 엔-버전 학습법의 구현을 위해 다수의 규칙 집합을 이용하여 최종 분류 결과를 도출해 내기 위한 기법과 습득된 규칙 합들로부터 분류 시스템을 구축하기 위한 최적의 규칙 집합의 조합을 찾기 위한 기법을 제시하고 다수의 사례 집합을 이용하여 엔-버전 학습법이 델보의 학습 환경에 미치는 영향을 평가하였다.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2004년도 심포지엄 논문집 정보 및 제어부문
• /
• pp.126-128
• /
• 2004

As e-mail is becoming increasingly important in every day life activity, mail users spend more and more time organizing and classifying the e-mails they receive into folder. Many existing recommendation systems or text classification are mostly focused on recommending the products for the commercial purposes or web documents. So this study aims to apply these application to e-mail more necessary to users. This paper suggests a dynamic recommendation agent system based on Rule Filtering Component recommending the relevant category to enable users directly to manage the optimum classification when a new e-mail is received as the effective method for E-Mail Management. Moreover we try to improve the accuracy as eliminating the limits of misclassification that can be key in classifying e-mails by category. While the existing Bayesian Learning Algorithm mostly uses the fixed threshold, we prove to improve the satisfaction of users as increasing the accuracy by changing the fixed threshold to the dynamic threshold. We designed main modules by rule filtering component for enhanced scalability and reusability of our system.

• 한국멀티미디어학회논문지
• /
• 제20권6호
• /
• pp.927-934
• /
• 2017

As the technology of drone develops, the use of drone is increasing, In addition, the types of sensors that are inside of smart phones are becoming various and the accuracy is enhancing day by day. Various of researches are being progressed. Therefore, we need to control drone by using smart phone's sensors. In this paper, we propose the most suitable machine learning model that matches the gyro sensor data with drone's moving. First, we classified drone by it's moving of the gyro sensor value of 4 and 8 degree of freedom. After that, we made it to study machine learning. For the method of machine learning, we applied the One-Rule, Neural Network, Decision Tree, and Navie Bayesian. According to the result of experiment that we designated the value from gyro sensor as the attribute, we had the 97.3 percent of highest accuracy that came out from Naive Bayesian method using 2 attributes in 4 degree of freedom. On and the same, in 8 degree of freedom, Naive Bayesian method using 2 attributes showed the highest accuracy of 93.1 percent.