• The Transactions of the Korea Information Processing Society
• /
• v.7 no.1
• /
• pp.286-295
• /
• 2000

We present a novel statistically adaptive method using the Minimum Description Length(MDL) principle for unsupervised segmentation of magnetic resonance(MR) images. In the method, Markov random filed(MRF) modeling of tissue region accounts for random noise. Intensity measurements on the local region defined by a window are modeled by a finite Gaussian mixture, which accounts for image inhomogeneities. The segmentation algorithm is based on an iterative conditional modes(ICM) algorithm, approximately finds maximum ${\alpha}$ posteriori(MAP) estimation, and estimates model parameters on the local region. The size of the window for parameter estimation and segmentation is estimated from the image using the MDL principle. In the experiments, the technique well reflected image characteristic of the local region and showed better results than conventional methods in segmentation of MR images with inhomogeneities, especially.

• Proceedings of the Korean Information Science Society Conference
• /
• 2006.06a
• /
• pp.49-51
• /
• 2006

분자 wDNF기계를 통해 해 집단을 병렬적으로 탐색하여 유망한 텀들을 선택한 후 그를 구성하는 변수들의 분포를 평가, 확률 모델을 확립하고 그로부터 다음 세대의 해 집단을 구성함으로써 진화 알고리즘의 확장인 EDA을 DNA컴퓨팅으로 모델링한다. 또한 희박한(sparse) 해 집단에서 간략한 (parsimonious) wDNF모델을 항께 찾으므로 단순히 해 집단의 분포만을 진화시켜 나가는 것이 아니라 모델의 구조도 같이 최적화 시켜 나가는 방안을 제시한다.

• Communications for Statistical Applications and Methods
• /
• v.11 no.3
• /
• pp.465-474
• /
• 2004

Data visualization is the automation process and the discovery process to data sets in an effort to discover underlying information from the data. It provides rich visual depictions of the data. It has distinct advantages over traditional data analysis techniques such as exploring the structure of large scale data set both in the sense of number of observations and the number of variables by allowing great interaction with the data and end-user. We discuss the principles of data visualization and evaluate the characteristics of various tools of visualization according to these principles.

• Proceedings of the Korean Information Science Society Conference
• /
• 2006.10b
• /
• pp.412-415
• /
• 2006

본 논문에서는 파라미터에 대한 정보가 없는 데이터, 즉, 각각의 이벤트 발생에 불확실성이 존재하는 데이터들에 대한 인과 관계의 학습을 위해 그래픽 모델인 베이지안 네트워크를 사용하였다. 이를 위해 기존에는 주로 네트워크 학습에 K2, Sparse Candidate 등의 방법이 사용되었다. 학습 및 추론에 있어서 어떻게 하면 기존의 방법보다 정확하고 빠르게 처리할 수 있을지에 대한 개선된 알고리즘을 제시하고 다른 알고리즘들과의 성능 비교를 통해 제시한 방법론이 보다 좋은 성능을 가짐을 보였다.

• Journal of KIISE:Software and Applications
• /
• v.27 no.5
• /
• pp.521-528
• /
• 2000

One of the hidden problems in a domain theory is that an agent does not understand the meaning of its action. Graphplan uses mutex to improve efficiency, but it does not understand negation and suffers from a redundancy problem. Introducing a negative function not in IPP partially helps to solve this kind of problem. However, using a negative function comes with its own price in terms of time and space. Observing that a human utilizes opposite concept to negate a fact based on MDL principle, we hypothesize that using a positive atom rather than a negative function to represent a negative fact is a more efficient technique for building an intelligent agent. We show empirical results supporting our hypothesis in IPP domains. To autonomously learn the human-like concept, we generate a cycle composed of opposite operators from a domain theory and extract opposite literals through experimenting with the operators.