• Journal of the Korean Institute of Intelligent Systems
• /
• v.11 no.9
• /
• pp.849-854
• /
• 2001

GTM(Generative Topographic Mapping) model is a probabilistic version of the SOM(Self Organizing Maps) which was proposed by T. Kohonen. The GTM is modelled by latent or hidden variables of probability distribution of data. It is a unique characteristic not implemented in SOM model, and, therefore, it is possible with GTM to analyze data accurately, thereby overcoming the limits of SOM. In the present investigation we proposed a BGTM(Bayesian GTM) combined with Bayesian learning and GTM model that has a small mis-classification ratio. By combining fast calculation ability and probabilistic distribution of data of GTM with correct reasoning based on Bayesian model, the BGTM model provided improved results, compared with existing models.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2001.12a
• /
• pp.287-290
• /
• 2001

Bishop이 제안한 generative Topographic Mapping(GTM)은 Kohonen이 제안한 자율 학습 신경망인 Self Organizing Maps(SOM)의 확률적 버전이다. 본 논문에서는 이러한 GTM 모형에 베이지안 추론을 결합하여 작은 오분류율을 가지는 분류 알고리즘인 베이지안 GTM(Bayesian GTM)을 제안한다. 이 방법은 기존의 GTM의 빠른 계산 처리 능력과 베이지안 추론을 이용하여 기존의 분류 알고리즘보다 우수한 결과가 나타남을 실험을 통하여 확인하였다.

• Proceedings of the Korea Institute of Convergence Signal Processing
• /
• 2003.06a
• /
• pp.114-117
• /
• 2003

본 논문에서는 video stream내의 움직이는 객체 정보를 추정하고 동적 GTM(genetic tree-map) 알고리즘을 사용하여 얼굴 영역 검출 기법을 제안한다. 기존의 일반적인 객체 추정 기법은 클러스터(cluster)과정을 통하여 영상 정보를 분할하고 그 중 움직이는 객체 부분을 복원함으로서 추정하였다. 제안하는 기법은 BMA(block matching algorithm)［1］ 알고리즘을 사용하여 video stream 에서 움직이는 객체 정보를 얻고 클러스터 알고리즘으로 PCA(principal component analysis)를 사용한다. PCA 기법은 입력 데이터에 관해 통계적 특성을 이용하여 주성분을 찾는다. 주축과 영역분할 알고리즘을 사용하여 데이터를 분할하고, 분할된 객체 정보를 사용하여 특정 객체만을 추정하는 것이 가능하다. 이렇게 추정된 객체를 얼굴영역의 feature에 대하여 신경망 학습인 동적 GTM 알고리즘을 사용하여 생성된 동적 GTM 맵의 정보에 따라 객체의 얼굴영역만을 추출해 낼 수 있다［2-6］.

• Journal of Nuclear Fuel Cycle and Waste Technology(JNFCWT)
• /
• v.12 no.2
• /
• pp.135-151
• /
• 2014

It is very important to properly understand such "Transport Pathways" elements as "Pipe" and "Cell" pathways in commercial GoldSim Transport Module (GTM) for developing higer quality models and programs for performance assessment of complex radioactive waste repositories. With an illustrative case under an earthquake scenario, by which an increasement in the groundwater flow rate occurs though the geological medium, ways of avoiding possible modeling errors in the nuclide transport modeling in the radioactive waste repository system for its safety assessment by utilizing such pathways are discussed and a proper usage of the pathways is proposed.

• Journal of KIISE:Software and Applications
• /
• v.29 no.6
• /
• pp.386-395
• /
• 2002

In this paper, we suggest dynamic genetic tree-maps(DGTM) using optimal features on recognizing data. The DGTM uses the genetic algorithm about the importance of features rarely considerable on conventional neural networks and introduces GTM(genetic tree-maps) using tree structure according of the priority of features. Hence, we propose the extended formula, DGTM(dynamic GTM) has dynamic functions to separate and merge the neuron of neural network along the similarity of features.

• Journal of Broadcast Engineering
• /
• v.22 no.4
• /
• pp.496-508
• /
• 2017

In this paper, we propose a Tone Mapping Operator (TMO) that preserves global contrast and precisely preserves boundary information. In order to reconstruct a High Dynamic Range (HDR) image to a Low Dynamic Range (LDR) display by using Threshold value vs. Intensity value (TVI) based on Human Visual System (HVS) and contrast value. As a result, the global contrast of the image can be preserved. In addition, by combining the boundary information detected using Guided Image Filtering (GIF) and the detected boundary information using the spatial masking of the Just Noticeable Difference (JND) model, And improved the perceived image quality of the output image. The conventional TMOs are classified into Global Tone Mapping (GTM) and Local Tone Mapping (LTM). GTM preserves global contrast, has the advantages of simple implementation and fast execution time, but it has a disadvantage in that the boundary information of the image is lost and the regional contrast is not preserved. On the other hand, the LTM preserves the local contrast and boundary information of the image well, but some areas are expressed unnatural like the occurrence of the halo artifact phenomenon in the boundary region, and the calculation complexity is higher than that of GTM. In this paper, we propose TMO which preserves global contrast and combines the merits of GTM and LTM to preserve boundary information of images. Experimental results show that the proposed tone mapping technique has superior performance in terms of cognitive quality.

• Journal of Information Processing Systems
• /
• v.13 no.3
• /
• pp.546-558
• /
• 2017

The amount of multimedia traffic over the Internet has been increasing because of the development of networks and mobile devices. Accordingly, studies on multicast, which is used to provide efficient multimedia and video services, have been conducted. In particular, studies on centralized multicast tree construction have attracted attention with the advent of software-defined networking. Among the centralized multicast tree construction algorithms, the group Takahashi and Matsuyama (GTM) algorithm is the most commonly used in multiple multicast tree construction. However, the GTM algorithm considers only the network-cost overhead when constructing multicast trees; it does not consider the temporary service disruption that arises from a link change for users receiving an existing service. Therefore, in this study, we propose a multiple multicast tree construction algorithm that can reduce network cost while avoiding considerable degradation of service quality to users. This is accomplished by considering both network-cost and link-change overhead of users. Experimental results reveal that, compared to the GTM algorithm, the proposed algorithm significantly improves the user-experienced quality of service by substantially reducing the number of linkchanged users while only slightly adding to the network-cost overhead.

• Proceedings of the Korean Society for Bioinformatics Conference
• /
• 2002.06a
• /
• pp.49-61
• /
• 2002

현대적 실험방법 및 유전공학의 발전으로 최근 생물학적 자료는 비약적으로 늘어나고 있다. 이러한 자료의 기계학습을 이용한 분석방법은 많은 비용과 시간을 요구하는 전통적인 생물적 실험에 있어서 실험 시간을 단축시켜주고 실험비용을 줄여 주게 된다. 본 논문에서는 특별히 micro array data의 분석에 있어서 graphical model에 기반한 기계학습 방법들을 소개한다. 이중 GTM 은 특히 시각화 효과가 뛰어난 방법으로 Graphical model 에 기반한 GTM의 제반 특성을 소개하고 이를 yeast data의 분석에 적용시킨 결과를 자세히 알아보고자 한다. (**Presentation file을 수신 보관 중)

• Journal of Soil and Groundwater Environment
• /
• v.24 no.1
• /
• pp.10-16
• /
• 2019

Alkalinity is an essential parameter for understanding geochemical processes and calculating partial pressure of $CO_2$, dissolved inorganic carbon, and mineral saturation indices. The Gran Titration Method (GTM) is one of the most accurate methods for measuring the alkalinity in water samples. However, this method has not been widely employed in measuring groundwater alkalinity in Korea, probably due to inadequate and insufficient understanding of the method. In this regard, this article was prepared to introduce GTM and related know-hows learned from the authors' experiences in measuring alkalinity. This paper also introduces a MS Excel-based alkalinity calculator as a handy tool for GTM.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2005.11a
• /
• pp.1137-1140
• /
• 2005

기존의 방대한 무선 인프라와 최근 이동 단말기의 발달로 무선통신 기술을 이용한 서비스 이용이 잦아졌으며, 이를 관리하기 위해 기존의 모니터링 시스템들은 표준화되지 않은 지리 정보 속성을 바탕으로 단일 이동 단말기에서만 적용 되었다. 따라서 표준화된 GIS(Geographic Information System)에서 이동하는 여러 개의 무선 시스템을 효율적으로 관리하며 제어 할 수 있는 시스템이 필요하게 되었다. 본 논문은 WIPI(Wireless Internet Platform for Interoperability)에서 제공되는 GML(Geography Markup Language) 위치 정보 시스템을 서버에서 통합 추적 할 수 있는 모니터링 시스템을 설계한다. 또한 이 시스템은 OGC(Open GIS Consortium)에서 제안한 GML 3.1을 기반으로, 여러 WIPI 사용자들의 공간 정보를 획득 및 저장하여 이를 추적 관리 할 수 있다.