• Proceedings of the Korean Information Science Society Conference
• /
• 2004.04b
• /
• pp.595-597
• /
• 2004

본 연구에서는 베이지안망을 기초로 불임환자의 임상 데이터에 대한 다양한 실험을 전개한다. 실험을 통해 임신여부에 영향을 주는 요인들간의 상호 의존성을 분석하고. 또 제약조건이 다른 다양한 베이지안망의 대표적 유형으로 나이브 베이지안망(NBN), 베이지안망으로 확장한 나이브 베이지안망(BAN), 일반 베이지안앙(GBN) 분류기들의 분류성능을 서로 비교 분석한다. 베이지안망을 적응할 때 변수의 수가 많아짐에 따라 베이지안망의 구조를 학습하는데 탐색공간이 넓어져 시간의 요구량이 급격히 많아진다. 따라서 이런 탐색공간을 효율적으로 줄이기 위하여 클래스 노드의 Markov blanket에 속한 특징들로 집합을 축소하는 것을 제안하고, 실험을 통해 이 특징 축소 방법이 베이지안망 분류기들의 성능을 높여 줄 수 있는지 알아본다.

• The KIPS Transactions:PartB
• /
• v.9B no.5
• /
• pp.625-634
• /
• 2002

In this paper, we conducted various experiments with Bayesian networks in order to analyze clinical data of infertility patients. With these experiments, we tried to find out inter-dependencies among important factors playing the key role in clinical pregnancy, and to compare 3 different kinds of Bayesian network classifiers (including NBN, BAN, GBN) in terms of classification performance. As a result of experiments, we found the fact that the most important features playing the key role in clinical pregnancy (Clin) are indication (IND), stimulation, age of female partner (FA), number of ova (ICT), and use of Wallace (ETM), and then discovered inter-dependencies among these features. And we made sure that BAN and GBN, which are more general Bayesian network classifiers permitting inter-dependencies among features, show higher performance than NBN. By comparing Bayesian classifiers based on probabilistic representation and reasoning with other classifiers such as decision trees and k-nearest neighbor methods, we found that the former show higher performance than the latter due to inherent characteristics of clinical domain. finally, we suggested a feature reduction method in which all features except only some ones within Markov blanket of the class node are removed, and investigated by experiments whether such feature reduction can increase the performance of Bayesian classifiers.

• Management & Information Systems Review
• /
• v.38 no.3
• /
• pp.187-210
• /
• 2019

Companies shall hold social responsibility as a member of the social community. Corporate social responsibility uses corporate resources, yet it plays important roles in reducing social imbalance. Their responsibilities are highly associated with the corporate sustainability. Many earlier studies on the association between corporate social responsibility and corporate sustainability have been attempted. Yet it should be mentioned that they do not show a variety of realities as linearity between dependent variables and independent variables were assumed. Thus, this study aims to analyze Markov blanket, a node of minimum descriptive variables that relieve a rigid assumption among variables and affect corporate sustainability by using Bayesian network. Sensitivity analysis was used to elicit how other variables affect by reflecting the complex reality when real factors are changed. As an important result of this study, the firm's future earnings sustainability is naturally related to operating earnings, and as the corporate governance structure is sound, the firm is able to steadily fulfill its social responsibility. However, the fact that the size of a company is large does not mean that it is in good compliance with corporate laws. This would not be unrelated to the fact that many of today's companies are not complying with the law and are suffering social condemnation. Results from this study will serve as a useful analytic tool when investors and creditors showing interests in corporate sustainability for assessing the value of companies and making investment decisions. Moreover, they can be used as references for relevant agency supervising capital markets to establish or improve appropriate institutions aimed at improving corporate sustainability.

• Journal of Advanced Marine Engineering and Technology
• /
• v.40 no.5
• /
• pp.437-446
• /
• 2016

Feature selection has become an essential technique to reduce the dimensionality of data sets. Many features are frequently irrelevant or redundant for the classification tasks. The purpose of feature selection is to select relevant features and remove irrelevant and redundant features. Applications of the feature selection range from text processing, face recognition, bioinformatics, speaker verification, and medical diagnosis to financial domains. In this study, we focus on filter methods based on information entropy : IG (Information Gain), FCBF (Fast Correlation Based Filter), and mRMR (minimum Redundancy Maximum Relevance). FCBF has the advantage of reducing computational burden by eliminating the redundant features that satisfy the condition of approximate Markov blanket. However, FCBF considers only the relevance between the feature and the class in order to select the best features, thus failing to take into consideration the interaction between features. In this paper, we propose an improved FCBF to overcome this shortcoming. We also perform a comparative study to evaluate the performance of the proposed method.

• Management & Information Systems Review
• /
• v.36 no.1
• /
• pp.1-19
• /
• 2017

One of the methods of securing the reliability of accounting information is maintaining high audit quality. The first step of improving audit quality is lowering audit engagement risks. Thus, this study analyzed the relationship between the characteristics of accounting firms and audit engagement risks based on the Bayesian Network. For this, Markov Blanket, the minimum explanatory variable set, which affects audit engagement risks, was presented, and based on the drawn causal relationship, sensitivity analysis was conducted to verify the characteristics of accounting firms, which affect audit engagement risks. The existing preceding research that used multiple regression analysis presumes the linearity between explanatory variables and dependent variables, so there was a limit in drawing the relationship between explanatory variables. Therefore, this study figured out the interdependence between variables using the General Bayesian Network and examined the impact that each variable has finally on audit engagement risks that affects the audit quality. The results of this study would greatly contribute to improving the efficiency of the supervisory task by allowing a supervisory institution to identify an accounting firms that does not manage audit engagement risks properly and to improve the supervision of the accounting firms in advance. In addition, this study will be used as a reference when a supervisory institution would improve the system related to audit quality by presenting the characteristics of accounting firms related to the audit quality.