• Journal of the military operations research society of Korea
• /
• v.36 no.3
• /
• pp.43-56
• /
• 2010

Recently, the acquisition environment of the Korean defense weapon system is emphasizing more the importance of cost analysis in terms of efficient execution for defense acquisition budget. While cost analysis, however, is emphasized in law and process, its infrastructures are still insufficient We have been using computerized cost models to obtain an estimate at early phase of project. But those models have been developed by foreign companies, and so they have many limitations when using in Korean defense environment. For this reason, it began to sympathize that we need the development of the Korean version cost estimation model suitable for our defense industry environment, and now many studies are proceeding. In this study, we suggest Cost Estimating Relationships(CERs) developing methodologies which is key logics of Korean version cost estimation model. Especially, we proposed a new CER's development process depending upon data characteristics such as, multicolinearity, outlier, small samples and heteroscedasticity. Also, we presented a case study for artillery weapon system using these methods we developed. We find that these CERs could be verified through theoretical methods.

• Journal of Navigation and Port Research
• /
• v.42 no.3
• /
• pp.187-194
• /
• 2018

Aids to Navigation (AtoN) is a marine traffic safety facility used to facilitate the safe and efficient movement of shipping and enhance the protection of the marine environment by the regulations or guidelines of The International Association of Marine Aids to Navigation and Lighthouse Authorities (IALA). Our country is managing AtoN to provide consistent services for AtoN users, although an average of 141 AtoN accidents occur annually. An AtoN accident forces non-planned work on the managers to resort the function, and causes psychological anxiety for its users, ultimately resulting in economic losses. This study developed the Social Cost evaluation model of AtoN accidents. The model can be used to quantify the manager's economic activities related to the shutdown and recovery, as well as the cost associated with the inconvenience to AtoN users. The Social Cost evaluation model of AtoN accidents is proposed as the sum of the encounter cost, administration cost and risk cost.

• Journal of Navigation and Port Research
• /
• v.43 no.3
• /
• pp.166-171
• /
• 2019

Aids to Navigation(AtoN) Accident can reduce the credibility of AtoN by adding extra works to the administrators, and cause burdens and inconveniences to the users during their navigation leading to social costs (or economic losses). Prior research on social cost estimation model to determine the Encounter cost, Administrative cost and Risk cost on the Aton model was done to estimate the economic losses due to AtoN Accident. However, this research could not provide adequate study results on risk cost estimates. Therefore, this study complements the estimation model presented in the previous study, and the estimated social cost of 249 AtoN Accidents that occurred during the last 10 years (2008-2017) in the Busan Regional Office of Oceans and Fisheries jurisdiction area. The risk cost was estimated using contingent valuation method since the AtoN is a non-market goods. As a result, the social cost of AtoN Accidents for the past 10 years was estimated to total 12.4 billion won including Encounter cost of 3.1 billion won (25.38%), Administrative cost of 1.7 billion won (13.62%), and Risk cost of 7.5 billion won (61.01%).

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2007.12a
• /
• pp.241-242
• /
• 2007

This paper aims to determination of container ship's economical replacement policy by Fuzzy Annual Equivalent Cost Method Especially, though there was much vagueness to expense for calculation of container ship's economical life, this paper expressed such vague nature with a fuzzy number. And this paper developed the fuzzy expense model with a fuzzy number, and to analyze more practically than an expense model, was decided economical life-age with an application in a containership of various volumes.

• The Transactions of the Korea Information Processing Society
• /
• v.1 no.2
• /
• pp.194-201
• /
• 1994

Recently, as application fields of software is extended, relative importance of software make a gradual increase and importance of development cost is being increased. However, as former evaluation model of development cost evaluate at the functional point of view for the most part, at this paper, I intend to propose evaluation model of software development cost COSMOS-Q(COSt MOdel for Subcontract-Quality) which one evaluate also in quality as well as function. The model proposed in this paper set the goal at software orderer evaluate software cost exactly with only order specification information. At this paper, I proposed cost evaluation model and evaluated it's validity refering to review result in ISO/SC7 about a software quality feature with extraction of quality feature factor which produce change of cost and set up the evaluation measure adoptable as order condition.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.18 no.11
• /
• pp.2664-2669
• /
• 2014

Recently, many research efforts have paid attention to spectrum pooling meshanism that provides efficient way to manage transient spectrum requests of secondary users in cognitive radio networks. Spectrum pooling is maintained by WSP (wireless service provider). WSP leases the spectrums stored in her spectrum pool to secondary users with being paid for it. It is natural that WSP tries to get profits as much as possible, which implies the WSP tries to minimize the cost required for maintaining her spectrum pool. In this paper, we model the spectrum pool into a probabilistic inventory model. Assuming secondary users' spectrum requests follow normal distribution, we give a strategic way that minimizes the maintenance cost of the spectrum pool. By a series of simulations, we show that WSP can reduce the total maintenance cost through our inventory model-based spectrum pooling than maintaining fixed inventory level.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2000.04a
• /
• pp.526-529
• /
• 2000

본 연구는 물류 개선이 시급하고, 전자상거래 등으로 인한 물류 거점의 확보가 중요시되고 있는 현실의 문제점을 개선하기 위해 Simulated Annealing 알고리즘을 제안한다. 이미 위치가 설정된 노드의 요구에 최소비용으로 원활히 대응하기 위한 다 물류센터의 최적 위치와 할당영역, 규모를 결정하고, 각 물류센터가 담당한 영역의 노드를 반드시 한번씩 경유하여 처음 위치로 돌아오는 최소비용의 경로를 탐색한다. 이러한 과정을 위해 두 개의 모델인 1) 다 물류센터의 위치결정을 통한 규모결정과 영역할당모델, 2) 경로계획 모델을 개발한다. 거리산정방법과 물동량에 따른 가중치를 적용한 다양한 물류센터의 위치결정과 규모결정, 각 물류센터별 영역을 할당하여 단 물류센터의 문제로 만든 후 경로계획을 실시하여 최소비용의 경로(Tour)를 구하는 알고리즘을 개발하고 Sample을 적용한다.

• 한국IT서비스학회:학술대회논문집
• /
• 2003.05a
• /
• pp.145-152
• /
• 2003

본 연구의 선행 연구에서는 증권시스템의 기능점수를 직접 구하여 기능점수와 운영비용과의 회귀분석을 실시하였다. 수집된 자료의 건수가 적었던 관계로 통계적 유의성을 충분하게 확보하지 못하였다. 따라서 본 연구에서는 증권시스템의 기능점수를 직접 측정하는 것이 현실적으로 많은 제약이 있음을 감안하여, 비교적 자료 수집이 용이한 프로그램 본 수를 측정하였다. 이러한 프로그램 본 수는 스텝 수로 1차 변환이 되었고, 스텝 수는 다시 기능점수로 2차 변환이 되었다. 이렇게 변환된 기능점수와 운영비용과의 회귀분석을 실시하였으며, 증권정보시스템 운영비용 추정 모델을 제시하였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.11a
• /
• pp.712-715
• /
• 2023

딥러닝 기술의 발전은 고품질의 대규모 데이터에 크게 의존한다. 그러나, 데이터의 품질과 일관성을 유지하는 것은 상당한 비용과 시간이 소요된다. 이러한 문제를 해결하기 위해 최근 연구에서 최소한의 비용으로 최대의 성능을 추구하는 액티브 러닝(active learning) 기법이 주목받고 있는데, 액티브 러닝은 모델 관점에서 불확실성(uncertainty)이 높은 데이터들을 샘플링 하는데 중점을 둔다. 하지만, 레이블 생성에 있어서 여전히 많은 시간적, 자원적 비용이 불가피한 점을 고려할 때 보완이 불가피 하다. 본 논문에서는 의사-라벨링(pseudo labeling)을 활용한 준지도학습(semi-supervised learning) 방식과 학습 손실을 동시에 사용하여 모델의 불확실성(uncertainty)을 측정하는 방법론을 제안한다. 제안 방식은 레이블의 신뢰도(confidence)와 학습 손실의 최적화를 통해 비용 효율적인 데이터 레이블 생성 방식을 제안한다. 특히, 레이블 데이터의 품질(quality) 및 일관성(consistency) 측면에서 딥러닝 모델의 정확도 성능을 높임과 동시에 적은 데이터만으로도 효과적인 학습이 가능할 수 있는 메커니즘을 제안한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.11a
• /
• pp.402-405
• /
• 2023

본 연구의 목적은 발생한 거래에 대해 적합한 회계 계정을 선택하는 예측 모델을 생성하는 것이다. 이를 통해 단기적으로 보조 수단으로 활용하여 회계 전표 승인에 대한 소요 시간을 단축하거나, 장기적으로 손익계산서가 일정 오차율 이내에서 자동으로 빠르게 작성됨으로써 재무 정보의 적시성을 올려주고, 기업의 실적을 나타냄에 있어 개별 담당자의 주관이 개입될 여지가 줄어든다는 면에서 재무 정보의 객관성을 올려줄 것으로 기대한다. 제안하는 모델은 비용 전표 이력의 적요를 자연어 처리하고 학습한 모델을 통해 1 차(공시용) 비용 계정을 분류한다. 분류 결과를 범위로 활용하는 기계 학습을 활용하여 좀 더 세밀한 범위의 2 차(관리용) 비용 계정을 분류하였다.