• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2005년도 Proceedings of ISRS 2005
• /
• pp.620-623
• /
• 2005

KARI has prepared Calibration and Validation activities for the KOMPSAT-2 (KOrea Multi-Purpose SATellite-2) MSC data that will be launched at the end of this year. Firstly, we divided the Cal/Val activities of it to four parts, Spatial, Spectral, Radiometric and Geometric, and defmed the detailed Cal/Val items from them. Secondly, Cal/Val targets have been defined and manufactured for the role of them. Thirdly, we have made the plan and the procedure for the Cal/Val items, developed the codes for them, studied more detailed method to do them, and trained the Cal/Val activities using the foreign satellite image data by ourselves. KARI has been now setting up the KOMPSAT-2 LEOP plan with the Cal/Val activities, and probably will finish the EOP Cal/Val activities for the KOMPSAT-2 MSC data by the next April or May.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2023년도 추계학술발표대회
• /
• pp.311-313
• /
• 2023

딥러닝에 대한 관심이 증가함에 따라 다양한 분야의 연구자 사이에 딥러닝 모델의 적용 및 재현이 중요한 작업으로 자리잡았다. 하지만 모델을 재현하고 활용하는데 있어 다양한 환경과 자원의 한계가 발생하여 문제가 되고 있다. 이러한 문제를 해결하기 위해 본 논문에서는 국가연구데이터커먼즈체계인 KRDC 프레임워크를 활용하여 딥러닝 학습 모델의 재현 방안을 제안하였다. 이를 통해 딥러닝 연구에 익숙하지 않은 사용자도 학습 모델의 적용 및 활용을 용이하게 할 수 있음을 확인하였다. KRDC 프레임워크는 사용자가 원하는 데이터와 태스크를 정의하고, 워크플로우로 구성, 학습 모델의 재현 및 활용을 지원한다.

• 대한임상검사과학회지
• /
• 제36권1호
• /
• pp.19-26
• /
• 2004

According to increase of domestic blood components use, the quality control of blood components is necessary to support good products. The purpose of this study is used to provide the producing index of the good product as compared with the accuracy and validity for the distribution of the quality control data. The value of mean, standard deviation, 95% confidence interval and degree of normal distribution of data were calculated by univariate procedure, the value of monthly mean of each blood centers per items were compared by Analysis of Variance(ANOVA) test for the degree of distribution. When there was difference among the mean values, the Duncan's multiple range test was done to confirm the difference. Finally, methods for accessing accuracy and validity of the quality data was done by the Contingency table test. The quality data of five blood centers was showed to the normal distribution and it was in a acceptable range. For each blood centers, the monthly means of Hematocrit(Hct), Platelet(PLT) and pH were not significantly different except Hct of C center, PLT of B, D center and pH of A center. The quality data per items was graded according to quality to six level. As a result of the comparative analysis, the monthly means of Hct of C and E center was significantly different higher than that of D, B and A center. The monthly means of PLT of A center and pH of C center was significantly different higher than that of the others. In the accuracy and validity of the quality control data, C center for Hct, A center for PLT and C center for pH were better than the other. The C blood center was most satisfiable and stable in the quality control for blood component. If the quality control method used in C blood center is adopted in other blood centers, the prepared level of the blood component of the center will be improved partly.

• Genomics & Informatics
• /
• 제8권4호
• /
• pp.165-169
• /
• 2010

The Korean Bioinformation Center (KOBIC) is a national bioinformatics research center in Korea. We developed many bioinformatics algorithms and applications to facilitate the biological interpretation of OMICS data. Here we present an introduction to major bioinformatics resources of databases and tools developed at KOBIC. These resources are classified into three main fields: genome, proteome, and literature. In the genomic resources, we constructed several pipelines for next generation sequencing (NGS) data processing and developed analysis algorithms and web-based database servers including miRGator, ESTpass, and CleanEST. We also built integrated databases and servers for microarray expression data such as MDCDP. As for the proteome data, VnD database, WDAC, Localizome, and CHARMM_HM web servers are available for various purposes. We constructed IntoPub server and Patome database in the literature field. We continue constructing and maintaining the bioinformatics infrastructure and developing algorithms.

• 한국물환경학회지
• /
• 제34권5호
• /
• pp.462-469
• /
• 2018

There are differences in method and cycle of flow rate survey depending on purpose of the operating department. To verify and use results of flow data according to the purpose, flow data of the directly measured and tele monitoring system were compared to verify validity. Flow measurement in the Ministry of Environment is aimed at setting up a standard flow of target water quality for water quality management and securing flow data of low and normal water level seasons for water quality evaluation. In this study, correlation analysis result ($R^2$) of same time zone data by direct measurement and tele monitoring system (TMS) at Seombon D point, a unit watershed of Seomjin river, for six years ('10 ~ '15) according to implementation of Total Daily Maximum Load (TDML) was wading 0.716, boating 0.962 and on bridge 0.943, and effectiveness of measurement method was verified by characteristics of flow duration as a season of dry and low-water; normal and high water are appropriate for wading, boating, and on bridge respectively. Results revealed it is reasonable to use directly measured results using the wading and boating method for low water level and dry seasons, and TMS data for rainy seasons. It can be used important data for future policy decisions.

• 한국자기학회:학술대회 개요집
• /
• 한국자기학회 2008년도 Asian Magnetics Conference
• /
• pp.294.1-294.1
• /
• 2008

• 산업공학
• /
• 제18권4호
• /
• pp.418-430
• /
• 2005

We establish an efficient strategy for construction and operation of data consolidation center for multi-organization in public sector. First, we introduce important concepts on data consolidation center in public sector, and draw some success factors by analyzing several foreign and domestic cases. Second, we construct all the possible logical operational models of the center and investigate the properties and feasibility of the models. Third, we suggest a virtual operational environment for the two representative models selected by feasibility criteria among the possible logical models, and compare the two models in terms of operational cost. We also utilize AHP methodology to evaluate qualitative opinions on the two models from several experts in public information systems. As a result, we find the best alternative is the case in which all infrastructure and facilities for the center are provided by government, and common essential IT operations are integrated, associated data are consolidated and the whole operational work are outsourced to specialized IT operations service providers.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2002년도 ICCAS
• /
• pp.53.2-53
• /
• 2002

NARO Space Center is being developed as a national project for the Korea Space Development Program. Among the major missions of the Space Center, the Range Control System is the focal point for all command and control operation of the Space Center. The acquired data from the Tracking Stations and the on-site facilities is processed and distributed in the Control Center. Data processing or data fusion is needed for the exact tracking of the Launch Vehicle from several tracking systems. The first phase, which is the best telemetry source is selected among data streams that are received from each telemetry stations using some pre-defined criterion. Trajectory data and major telemetry parameters...

• 정보통신설비학회논문지
• /
• 제9권3호
• /
• pp.107-111
• /
• 2010

This paper is concerned with data center remote management service for demanding forecasting and reduction of energy usage. More particularly, intelligent server rack, mounted on inside of the data center, collects information about energy usage and temperature per server. Using this information, management platform forecasts energy demand in the future and automatically makes report according green environment raw. By providing the remote management service through remote terminals, users are not tied to a time and place to control device inside the data center. In this way, the data center remote management service enhances operability of the facility.

• 한국지리정보학회지
• /
• 제22권4호
• /
• pp.215-228
• /
• 2019

해양관측 정지궤도 위성인 GOCI(Geostationary Ocean Color Imager) 데이터는 대용량 산출물을 효과적으로 저장, 배포하기 위해 HDF5 자료 형식을 사용하고 있다. 해양위성센터에서는 HDF5(Hierarchical Data Format version5) 포맷에 익숙지 않은 일반 사용자를 위해 GDPS(GOCI Data Processing System)를 개발하여 관측자료와 함께 제공하고 있다. 그럼에도 불구하고 위성데이터 특성에 대한 이해와 GDPS의 사용법을 익혀야 하는 점, 그리고 위치정보와 속성정보가 분리되어 있는 HDF5 형식의 자료를 병합하고 가공하는 일은 쉽지 않은 일이다. 따라서 본 연구에서는 오픈소스 R과 rhdf5, data.table, matrixStats 패키지를 이용하여 GDPS를 이용하는 과정 없이도 HDF5 형식의 위성데이터를 손쉽게 활용할 수 있는 알고리즘을 개발하였다.