• 지능정보연구
• /
• 제14권4호
• /
• pp.1-18
• /
• 2008

사례 기반 계획 시스템은 과거의 유사한 사례 계획들을 이용함으로써 새로운 문제를 위한 계획을 효율적으로 생성 할 수 있다. 하지만 대부분의 기존 사례 기반 계획 시스템들은 사례 검색 및 사례 일반화를 위한 제한적 기능들만을 제공할 뿐만 아니라, 계획 생성과정에 사용자의 참여를 허용하지 않는다. 본 논문에서 제안하는 JCBP 시스템은 효율적인 메모리 사용과 사례 검색을 위해 각 도메인의 동일한 작업 목표를 가진 사례들을 개별 사례 베이스로 그룹화하고, 이들에 대한 색인을 유지한다. 또 이 시스템은 문제모델로부터 자동으로 추출한 휴리스틱 지식을 사례 적응 단계에 이용하며, 목표 회귀를 통한 사례 일반화 기능도 제공한다. 또한 JCBP 시스템은 대화형 모드를 통한 혼합 주도 계획 생성 기능도 제공한다. 이와 같이 JCBP 시스템은 문제 해결을 위해 사용자의 기호와 지식을 이용함으로써 사용자의 요구를 더 잘 만족하는 해 계획을 생성할 수 있을 뿐 아니라 계획 생성의 복잡도도 줄일 수 있다.

• 한국측량학회지
• /
• 제19권2호
• /
• pp.199-208
• /
• 2001

우리나라의 수치지도제작은 축척에 따라 독립적으로 제작되어 활용되어 왔다. 따라서 다른 축척의 공간자료를 원할 경우 다시 제작해야 하기 때문에 막대한 시간과 경제적인 손실을 가져온다. 이러한 문제점을 해결하기 위한 방안으로 대축척 수치지도를 원시데이터로 하여 다양한 소축척 수치지도를 제작하기 위한 지도 일반화의 연구가 많이 진행되고 있다. 본 연구는 여러가지 일반화 처리 기법중 단순화에 의한 선형요소와 건물 레이어의 소축척 변환 특성을 연구하는데 그 목적이 있다. 이를 위하여 단순화 처리를 위한 알고리즘을 제시하였으며, 1:5,000수치지도로부터 1:25,000 수치지도로의 변환에 이를 적용, 수치지도의 단순화 특성을 분석하였다. 연구결과 면적분배 알고리즘이 Douglas-Peucker 알고리즘이나, Reumann-Witkam 알고리즘보다 건물의 최초면적과 형상을 잘 유지함을 알 수 있었다.

• 한국군사과학기술학회지
• /
• 제24권4호
• /
• pp.382-391
• /
• 2021

Unsupervised domain adaptation often gives impressive solutions to handle domain shift of data. Most of current approaches assume that unlabeled target data to train is abundant. This assumption is not always true in practices. To tackle this issue, we propose a general solution to solve the domain gap minimization problem without any target data. Our method consists of two regularization steps. The first step is a pixel regularization by arbitrary style transfer. Recently, some methods bring style transfer algorithms to domain adaptation and domain generalization process. They use style transfer algorithms to remove texture bias in source domain data. We also use style transfer algorithms for removing texture bias, but our method depends on neither domain adaptation nor domain generalization paradigm. The second regularization step is a feature regularization by feature alignment. Adding a feature alignment loss term to the model loss, the model learns domain invariant representation more efficiently. We evaluate our regularization methods from several experiments both on small dataset and large dataset. From the experiments, we show that our model can learn domain invariant representation as much as unsupervised domain adaptation methods.

• 대한수학회보
• /
• 제44권2호
• /
• pp.215-223
• /
• 2007

We in this note introduce the concept of NCI rings which is a generalization of NI rings. We study the basic structure of NCI rings, concentrating rings of bounded index of nilpotency and von Neumann regular rings. We also construct suitable examples to the situations raised naturally in the process.

• Korean Journal of Mathematics
• /
• 제23권1호
• /
• pp.37-46
• /
• 2015

We in this note introduce the concept of semi-IFP rings which is a generalization of IFP rings. We study the basic structure of semi-IFP rings, and construct suitable examples to the situations raised naturally in the process. We also show that the semi-IFP does not go up to polynomial rings.

• 한국통신학회논문지
• /
• 제26권8B호
• /
• pp.1078-1084
• /
• 2001

In this paper, fuzzy-ARTMAP neural network is applied for compensating the nonlinearity of satellite communication channel. The fuzzy-ARTMAP is made of using fuzzy logic and ART neural network. By a match tracking process with vigilance parameter, fuzzy ARTMAP neural network achieves a minimax learning rule that minimizes predictive error and maximizes generalization. Thus, the system automatically learns a minimal number of recognition categories, or hidden units, to meet accuracy criteria. Simulation studies are performed over satellite nonlinear channels. The performance of proposed fuzzy-ARTMAP equalizer is compared with MLP-basis equalizers.

• 한국전산구조공학회:학술대회논문집
• /
• 한국전산구조공학회 2002년도 가을 학술발표회 논문집
• /
• pp.271-278
• /
• 2002

Structural optimization using improved higer-order convex approximation is proposed in this paper. The proposed method is a generalization of the convex approximation method. The order of the approximation function for each constraint is automatically adjusted in the optimization process. And also the order of each design variable is differently adjusted. This self-adjusted capability makes the approximate constraint values conservative enough to maintain the optimum design point of the approximate problem in feasible region. The efficiency of proposed algorithm, compared with conventional algorithm is successfully demonstrated in the Three-bar Truss example.

• 한국시뮬레이션학회:학술대회논문집
• /
• 한국시뮬레이션학회 1999년도 춘계학술대회 논문집
• /
• pp.147-151
• /
• 1999

This paper extends and enhances the existing inverse kinematics technique using the concept of motion characteristic capturing. Motion characteristic capturing is not about measuring motion by tracking body points. Instead, it starts from pre-measured motion data, extracts the motion characteristics, and applies them in animating other bodies. The resulting motion resembles the originally measured one in spite of arbitrary dimensional differences between the bodies. Motion characteristics capturing is a new principle in kinematic motion generalization to process measurements and generate realistic animation of human being or other living creatures.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• 제22권4호
• /
• pp.289-294
• /
• 2018

In this work, a recently developed semi-analytic technique, so called the residual power series method, is generalized to process higher-dimensional linear and nonlinear partial differential equations. The solutions obtained takes a form of an infinite power series which can, in turn, be expressed in a closed exact form. The results reveal that the proposed generalization is very effective, convenient and simple. This is achieved by handling the (m+1)-dimensional Burgers equation.

• 대한의용생체공학회:의공학회지
• /
• 제42권4호
• /
• pp.159-166
• /
• 2021

In order to analyze cell images, accurate segmentation of each cell is indispensable. However, the reality is that accurate cell image segmentation is not easy due to various noises, dense cells, and inconsistent shape of cells. Therefore, in this paper, we propose an algorithm that combines marker-based watershed segmentation and ellipse fitting method for glioblastoma cell segmentation. In the proposed algorithm, in order to solve the over-segmentation problem of the existing watershed method, the marker-based watershed technique is primarily performed through "seeding using local minima". In addition, as a second process, the concave point search using ellipse fitting for final segmentation based on the connection line between the concave points has been performed. To evaluate the performance of the proposed algorithm, we compared three algorithms with other algorithms along with the calculation of segmentation accuracy, and we applied the algorithm to other cell image data to check the generalization and propose a solution.