• Journal of Families and Better Life
• /
• v.26 no.5
• /
• pp.333-354
• /
• 2008

This study aimed to investigate the role of domain-specific causal mechanism information and domain-general conditional probability in young children's causal reasoning on psychology and biology. Participants were 121 3-year-olds and 121 4-year-olds recruited from seven childcare centers in Seoul, Kyonggi Province, and Busan. After participants watched moving pictures on psychological and biological phenomena, they were asked to choose appropriate cause and justify their choices. Results of this study were as follows: First, young children made different inferences according to domain-specific causal mechanisms. Second, the developmental level of causal mechanisms has a gap between psychology and biology, and biological knowledge was proved to be separate from psychological knowledge during the preschool period. Third, young children's causal reasoning was different depending on the interaction effect of domain-specific mechanisms and domain-general conditional probability: children could make more inferences based on domain-specific causal mechanisms if conditional probability between domain-appropriate cause and effect was evident. To conclude, it can be inferred that the role of domain-specific causal mechanisms and domain-general conditional probability is not competitive but complementary in young children's causal reasoning.

• Journal of KIISE:Software and Applications
• /
• v.33 no.10
• /
• pp.877-885
• /
• 2006

Recently it is proposed that the Bayesian networks used as conversational agent for topic inference is useful but the Bayesian networks require much time to model, and the Bayesian networks also have to be modified when the scripts, the database for conversation, are added or modified and this hinders the scalability of the agent. This paper presents a method to improve the scalability of the agent by constructing the Bayesian network from scripts automatically. The proposed method is to model the structure of Bayesian networks hierarchically and to utilize Noisy-OR gate to form the conditional probability distribution table (CPT). Experimental results with ten subjects confirm the usefulness of the proposed method.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2007.06a
• /
• pp.241-247
• /
• 2007

자궁 경부 세포진 영상의 핵 추출을 위해서는 영상의 배경과 핵 그리고 세포질 영역의 구분이 중요하다. 또한 정상 세포핵과 암종 세포핵의 구분 및 인식을 위해서는 세포핵들의 형태학적 특징을 이용한 분류 기준을 세워야한다. 본 논문에서는 자궁 경부 세포진 영상에서 세포핵의 후보 영역과 핵을 추출하기 위해 현미경 400배율 확대 사진을 획득하는 과정에서 훼손된 컬러 영상을 복원하기 위한 방법으로 Lighting Compensation을 적용하여 영상을 보정한다. 그리고 배경 영역과 세포핵 영역을 구분하기 위해 영상의 R,G,B 영역의 히스토그램의 분포를 이용하여 배경을 제거한다. 배경이 제거된 영상을 그레이 영상으로 변환 한 후, 히스토그램 명암도의 값을 이용하여 세포핵 영역과 세포질을 분류하여 세포핵 영역을 추출한다. 그리고 Kapur 방법을 적용하여 세포핵 영역의 엔트로피 누적확률을 구한 후, 영상을 이진화 한다. Kapur 방법이 적용된 이진화 영상에서 세포핵 영역의 중심과 주위 화소를 비교하는 $3\times3$ 마스크를 적용하여 영상의 미세한 잡음을 제거 한 후, 8방향 윤곽선 추적 알고리즘을 적용하여 최종적으로 세포핵 영역을 추출한다. 추출된 세포핵의 영역을 분류 및 인식하는 과정으로 세포의 외각의 방향성 정보, 핵의 크기, 그리고 면적 비율의 특징을 이용하여 퍼지 소속 함수를 설계한 후, 소속 함수의 소속도를 구하고 퍼지 추론 규칙을 적용하여 자궁 경부 세포진 영상에서 정상 세포핵 및 암종 세포핵을 인식한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2016.04a
• /
• pp.588-590
• /
• 2016

사용자의 성별은 기본적이면서도 중요한 마케팅 데이터다. 그러나 최근에는 개인정보보호 강화 추세로, 회원가입 시 성별이나 나이 등의 세부 정보를 입력하지 않는 간편 가입이 많아졌다. 이러한 입력되지 않은 정보 추출을 위해 성별 예측 연구의 필요성이 증가되었다. 성별이 입력된 사용자의 정보를 바탕으로 성별이 입력되지 않은 사용자의 성별을 예측하는 기존 연구가 다양한 방법으로 진행되어왔고, 우수한 식별이 가능한 기법들은 이진분류기인 SVM을 기반으로 한 연구가 다수 존재한다. 그러나 SVM 알고리즘은 이진 분류만 가능하기 때문에 성별예측에 대한 정확률은 알 수가 없다. 성별예측의 정확률을 활용하면 부정확한 분류를 예방할 수 있으며 상품추천의 가중치로 사용 될 수 있다. 본 연구는 확률을 기반으로 하여 정확률을 추론 가능한 나이브 베이지안을 응용한다. 그리고 데이터 집합 사례를 균형있게 늘려주는 SMOTE기법을 이용해 클래스 불균형 문제를 개선했으며 또한 성별 예측의 특성에 맞게 노이즈를 제거하고, 성별 분류에 확정적인 아이템에 가중치를 적용했다. 더불어 제안 방법을 실제 데이터에 적용시켜 우수성을 입증하였다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.2
• /
• pp.402-409
• /
• 2007

In this paper recognized for trajectory using Condensation algorithm. In this pater used fuzzy controller for recognized trajectory using fuzzy reasoning. The fuzzy system tract to the three-dimensional space for raw and roll movement. The joint angle ${\theta}_1$ of the manipulator rotate from $0^{\circ}\;to\;360^{\circ}$, and the joint angle ${\theta}_2$ rotate from $0^{\circ}\;to\;180^{\circ}$. The moving object of velocity display for recognition without error using Condensation algorithm. The tracking system demonstrated the reliability of proposed algorithm through simulation against used trajectory.

• Journal of Educational Research in Mathematics
• /
• v.17 no.2
• /
• pp.179-199
• /
• 2007

The purpose of this study was to discover differences between mathematically gifted students (MGS) and non-gifted students (NGS) when making probability judgments. For this purpose, the following research questions were selected: 1. How do MGS differ from NGS when making probability judgments(answer correctness, answer confidence)? 2. When tackling probability problems, what effect do differences in probability judgment factors have? To solve these research questions, this study employed a survey and interview type investigation. A probability test program was developed to investigate the first research question, and the second research question was addressed by interviews regarding the Program. Analysis of collected data revealed the following results. First, both MGS and NGS justified their answers using six probability judgment factors: mathematical knowledge, use of logical reasoning, experience, phenomenon of chance, intuition, and problem understanding ability. Second, MGS produced more correct answers than NGS, and MGS also had higher confidence that answers were right. Third, in case of MGS, mathematical knowledge and logical reasoning usage were the main factors of probability judgment, but the main factors for NGS were use of logical reasoning, phenomenon of chance and intuition. From findings the following conclusions were obtained. First, MGS employ different factors from NGS when making probability judgments. This suggests that MGS may be more intellectual than NGS, because MGS could easily adopt probability subject matter, something not learnt until later in school, into their mathematical schemata. Second, probability learning could be taught earlier than the current elementary curriculum requires. Lastly, NGS need reassurance from educators that they can understand and accumulate mathematical reasoning.

• The Journal of Korean Association of Computer Education
• /
• v.12 no.4
• /
• pp.65-75
• /
• 2009

The result verification is necessary to support a guarantee for the correctness of the task results be executed by any unspecified resources in desktop grid environments. Typically, voting-based and trust-based result verification schemes have been used in the environments. However, these suffer from two potential problems: waste of resources due to redundant replicas of each task and increase in turnaround time due to the inability to deal with a dynamic changeable execution environment. To overcome these problems, we propose a fuzzy inference-based replication scheme which can adaptively determine the number of replicas per task by using both trusty degree and result return probability of resources. Therefore our proposal can reduce waste of resources by determining the number of replicas meeting with a dynamic execution environment of desktop grids, not to mention an enhancement of turnaround time for entire asks. Simulation results show that our scheme is superior to other ones in terms of turnaround time, the waste of resources, and the number of re-replications per task.

• The Korean Journal of Applied Statistics
• /
• v.19 no.3
• /
• pp.505-519
• /
• 2006

We consider zero-inflated count data, which is discrete count data but has too many zeroes compared to the Poisson distribution. Zero-inflated data can be found in various areas. Despite its increasing importance in practice, appropriate statistical inference on zero-inflated data is limited. Classical inference based on a large number theory does not fit unless the sample size is very large. And regular Poisson model shows lack of St due to many zeroes. To handle the difficulties, a mixture of distributions are considered for the zero-inflated data. Specifically, a mixture of a point mass at zero and a Poisson distribution is employed for the data. In addition, when there exist meaningful covariates selected to the response variable, loglinear link is used between the mean of the response and the covariates in the Poisson distribution part. We propose a Bayesian inference for the zero-inflated Poisson regression model by using a Markov Chain Monte Carlo method. We applied the proposed method to a Korean oral hygienic data and compared the inference results with other models. We found that the proposed method is superior in that it gives small parameter estimation error and more accurate predictions.

• Journal of Educational Research in Mathematics
• /
• v.25 no.3
• /
• pp.323-345
• /
• 2015

This study examined the types of abduction appeared in the exploration activities of 'law of large numbers' in order to figure out relation between statistical reasoning and abduction. When the classroom discourse of students was analyzed by Peirce's abduction, Eco's abduction type and Toulmin's argument pattern, students used overcoded abduction the most in the discourse of abduction. However, there composed a low percent of undercoded abduction leading to various thinking, and creative abduction used to make new principles or theories. By the CAS calculators used in the process of reasoning, students were provided with empirical context to understand the concept of abstract probability, through which they actively participated in the argumentation centered on the reasoning. As a result, it was found that not only to understand the abduction, but to build statistical context with tools in the learning of statistical reasoning is important.

• Journal of KIISE:Software and Applications
• /
• v.34 no.6
• /
• pp.549-561
• /
• 2007

Vision-based scene understanding is to infer and interpret the context of a scene based on the evidences by analyzing the images. A probabilistic approach using Bayesian networks is actively researched, which is favorable for modeling and inferencing cause-and-effects. However, it is difficult to gather meaningful evidences sufficiently and design the model by human because the real situations are dynamic and uncertain. In this paper, we propose a learning method of Bayesian network that reduces the computational complexity and enhances the accuracy by searching an efficient BN structure in spite of insufficient evidences and training data. This method represents the domain knowledge as ontology and builds an efficient hierarchical BN structure under constraint rules that come from the ontology. To evaluate the proposed method, we have collected 90 images in nine types of circumstances. The result of experiments indicates that the proposed method shows good performance in the uncertain environment in spite of few evidences and it takes less time to learn.