• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2000.11a
• /
• pp.480-483
• /
• 2000

공학의 많은 응용분야에서 큰 회소 행렬(Large Sparse Matrices)에 대한 가장 작거나 또는 가장 큰 고유치(Eigenvalues)들을 요구하게 되는데, 이때 많이 이용되는 것은 Krylov Subspace로의 Projection방법이다. 대칭 행렬에 대해서는 Lanczos방법을, 비대칭 행렬에 대해서는 Biorhtogonal Lanczos방법을 이용할 수 있다. 이러한 기존의 알고리즘들은 새롭게 제안되는 병렬처리 시스템에서 효과적이지 못하다. 많은 프로세서를 가지는 병렬처리 컴퓨터 중에서도 분산 기억장치 시스템(Distributed Memory System)에서는 프로세서들 사이의 Data Communication에 필요한 시간을 줄이도록 해야한다. 본 논문에서는 기존의 Lanczos 알고리즘을 수정함으로써, 알고리즘의 동기점(Synchronization Point)을 줄이고 병렬화를 위한 입상(Granularity)을 증가시켜서 MPP인 Cray T3E에서 Data Communication에 필요한 시간을 줄인다. 많은 프로세서를 사용하는 경우 수정된 알고리즘이 기존의 알고리즘에 비해 더 나은 speedup을 보여준다.

• Electronics and Telecommunications Trends
• /
• v.35 no.3
• /
• pp.55-65
• /
• 2020

With the development of nonvolatile memory technology, Intel has released the Optane datacenter persistent memory module (DCPMM) that can be deployed in the dual in-line memory module. The results of research and experiments on Optane DCPMMs are significantly different from the anticipated results in previous studies through emulation. The DCPMM can be used in two different modes, namely, memory mode (similar to volatile DRAM: Dynamic Random Access Memory) and app direct mode (similar to file storage). It has buffers in 256-byte granularity; this is four times the CPU (Central Processing Unit) cache line (i.e., 64 bytes). However, these properties are not easy to use correctly, and the incorrect use of these properties may result in performance degradation. Optane has the same characteristics of DRAM and storage devices. To take advantage of the performance characteristics of this device, operating systems and applications require new approaches. However, this change in computing environments will require a significant number of researches in the future.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 1997.10a
• /
• pp.166-169
• /
• 1997

For parallel processing, the compiler partitions a loaded program into a set of tasks and makes a schedule for the tasks that will minimize parallel processing time for the loaded program. Building an optimal schedule for a given set of partitioned tasks of a program has known to be NP-complete. In this paper we introduce a GA(Genetic Algorithm)-based scheduling method in which a chromosome consists of two parts of a string which decide the number and order of tasks on each processor. An additional computation is used for feasibility constraint in the chromosome. By granularity theory, a partitioned program is categorized into coarse-grain or fine-grain types. There exist good heuristic algorithms for coarse-grain type partitioning. We suggested another GA adaptive to the coarse-grain type partitioning. The infeasibility of chromosome is overcome by the encoding and operators. The number of processors are decided while the GA find the minimum parallel processing time.

• Proceedings of the Korean Information Science Society Conference
• /
• 2008.06a
• /
• pp.343-344
• /
• 2008

최근 시스템의 복잡도가 증자함에 따라 보안 취약점 문제가 더욱 많이 발생하고 있다. 이를 해결하기 위해 보안 패치가 배포되고 있지만, 시스템 서비스의 중단이 필요하고 패치 자체의 안정성이 검증되지 못해 패치의 적용이 늦어지는 문제가 발생한다. 우리는 이러한 문제를 해결하기 위해 업데이트성이 없는 커널을 위한 함수 단위 동적 업데이트 시스템인 DUNK를 설계 하였다. DUNK는 서비스 중단 없는 업데이트를 가능케 하고, 보안 기법인 MAFIA를 이용해 안전한 업데이트를 수행한다. MAFIA는 바이너리 패치 코드의 접근 행위를 분석함으로써 패치된 함수가 기존 함수의 접근 권한을 상속받도록 하고, 이를 검증하는 기술을 제공한다. 본 논문에서는 DUNK의 설계와 MAFIA의 알고리즘 및 수행에 대해 기술한다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2005.07b
• /
• pp.352-354
• /
• 2005

1990년대 후반에 들어서면서 않은 기업들이 기업의 생산성 향상 및 고객 서비스를 위해 IT투자를 확대해감에 따라 기업 내에서도 다수의 이질적인 시스템 도입을 추진하게 되었다. 그 결과 않은 기업 내에는 호환성이 없는 시스템이 산재해 업무추진에 어려움을 주고 있으며, 이러한 복잡하고 까다로운 분산화된 시스템을 통합하고자 업계에서는 많은 노력을 기울이게 되었다. 그 결과로 등장한 기술이 웹서비스이며 이것은 분산화된 시스템을 통합하는 기존의 통합기술들의 문제점을 해결하고 분산된 애플리케이션을 저비용${\cdot}$고효율로 통합하기 위한 대안이 되었다. 본 논문에서는 웹서비스기반 시스템 개발을 위한 웹서비스 개발 절차를 제시하고 비즈니스 도메인에서 요구사항 분석을 통해 생성된 유스케이스모델을 기반으로 하는 웹 서비스 식별 방법을 제안한다. 유스케이스를 기반으로 서비스를 식별하므로 서비스에 도메인의 비즈니스가 잘 반영되도록 할 수 있으며 적절한 입도(granularity)을 갖는 서비스를 식별할 수 있다.

• Proceedings of the KOSOMBE Conference
• /
• v.1995 no.11
• /
• pp.177-182
• /
• 1995

Since one property(i.e. coarseness, orientation, regularity, granularity etc.) of ultrasound liver images was not sufficiently enough to classify the characteristics of livers, we used the multi-feature vectors from ultrasound images to diagnose the liver disease. The proposed classifier, which uses the multi-feature vectors and Bayes decision rule, performed well for the classification of normal, fat and cirrhosis liver. In our simulation, we used the Battacharyya distance and Hotelling Trace Criterion to select the best multi-feature vectors for the classifier and obtained less classification errors than other methods using single feature vector.

• Proceedings of the KIEE Conference
• /
• 2000.11d
• /
• pp.672-674
• /
• 2000

There are various ways in classification methodologies of data mining such as neural networks but the result should be explicit and understandable and the classification rules be short and clear. Rough set theory is a effective technique in extracting knowledge from incomplete and inconsistent information and makes an offer classification and approximation by various attributes with effect. This paper discusses granularity of knowledge for reasoning of uncertain concepts by using generalized rough set approximations based on hierarchical granulation structure and uses hierarchical classification methodology that is more effective technique for classification by applying core to upper level. The consistency rules with minimal attributes is discovered and applied to classifying real data.

• The Journal of the Korean dental association
• /
• v.10 no.1
• /
• pp.29-33
• /
• 1972

The author has studied histochemically on the giant cells appeqring in the 10 giant cell lesions which had been collected from the biopsies at the deaprtment of oral pathology, college of dentistry, Seoul national university and dental sections of several hospitals in Seoul. The results obtained were as follows: 1. The giant cells varied in size had 2 to 10 nuclei without any mitotic figures. 2. The giant cells showed high granularity in cytoplasm, extreme irregularity in outline of nuclear and cytoplasmic membrane, and very large and prominent nucleoli. 3. The giant cells were reacted moderately by Periodic Acid-Schiffraction. 4. The chtoplams of giant cells were stained moderately of deeply with Pyronin, and their nuclei with Methyl green. 5. In the sections stained with Iron-Hematoxylin, the giant cells showed microvacules and fine granular type mitochondria aggregated close to nuclei but not phagocytosed materials.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.16 no.3
• /
• pp.157-162
• /
• 2016

In this paper, a yield prediction model based on relevance vector machine (RVM) and a granular computing model (quotient space theory) is presented. With a granular computing model, massive and complex meteorological data can be analyzed at different layers of different grain sizes, and new meteorological feature data sets can be formed in this way. In order to forecast the crop yield, a grey model is introduced to label the training sample data sets, which also can be used for computing the tendency yield. An RVM algorithm is introduced as the classification model for meteorological data mining. Experiments on data sets from the real world using this model show an advantage in terms of yield prediction compared with other models.

• Asia pacific journal of information systems
• /
• v.5 no.1
• /
• pp.219-235
• /
• 1995

Complexity of some problems often transcends the problem solving capability of human individuals. As such, we cannot but take a collaborative approach to those complex problems. Collaboration while using computing systems can take place through shared objects. However, current commercial DBMSs do not provide a satisfactory control for concurrent access to objects shared by collaborators. A new concurrency control scheme is proposed which will help a group of people work in a more collaborative and natural way. The idea of softlock is refined into three different kinds of softlocks and the roles of collaborators are taken into account in the proposal of the new scheme. Although these softlocks are illustrated together with read/write hardlocks in this paper, the use of softlocks can be extended to be exploited with other kinds of hardlocks, for example, of granularity locking.