• The Korean Journal of Applied Statistics
• /
• v.16 no.2
• /
• pp.305-319
• /
• 2003

This paper provides a non-Bayesian method based on the expanded EM algorithm for segmenting the magnetic resonance images degraded by bias field. For the images with the intensity as a pixel value, many segmentation methods often fail to segment it because of the bias field(with low frequency) as well as noise(with high frequency). Our contextual approach is appropriately designed by using normal mixture model incorporated with Markov random field for noise-corrective segmentation and by using the penalized likelihood to estimate bias field for efficient bias filed-correction.

• Journal of the Korean Data and Information Science Society
• /
• v.12 no.2
• /
• pp.1-10
• /
• 2001

Neural networks we increasingly being seen as an addition to the statistics toolkit which should be considered alongside both classical and modern statistical methods. Neural networks are usually useful for classification and function estimation. In this paper we concentrate on function estimation using neural networks with weight decay factor The use of weight decay seems both to help the optimization process and to avoid overfitting. In this type of neural networks, the problem to decide the number of hidden nodes, weight decay parameter and iteration number of learning is very important. It is called the optimization of weight decay neural networks. In this paper we propose a automatic optimization based on genetic algorithms. Moreover, we compare the weight decay neural network automatically learned according to automatic optimization with ordinary neural network, projection pursuit regression and support vector machines.

• Journal of Conservation Science
• /
• v.16 s.16
• /
• pp.15-20
• /
• 2004

Two pieces of choson-Tongue(朝鮮通寶) minted at 1423 A.D. were analyzed by atomic absorption spectroscopy and neutron activation analysis. The measurement of lead isotope ratios was also carried out in order to predict the provenance of raw materials used for minting. The Chosen-Tongue was minted as bronze having the chemical compositions of $Cu\;90\%,\;Pb\;3\~4\%,\;Sn\;2\~3\%$, which were different from the typical composition of Chinese and Japanese coins. The results of lead isotope ratios showed that the provenance of raw materials used for minting had a possibility to be originated from South China. And application of statistical linear discriminant analysis (SLDA) to the provenance of lead used for minting of Chosen-Tongue was confirmed.

• Journal of Navigation and Port Research
• /
• v.31 no.1 s.117
• /
• pp.21-27
• /
• 2007

The wave load and its influence on the response of offshore structure have been well investigated through the statistical approach based on the linear theory. The linear approach has a limitation to apply the extreme condition such as extreme wave, which corresponds to extreme value of wave spectrum. The main topic of present study is to develop an efficient numerical method to predict wave load induced by extreme wave. As a numerical method, finite element method based on variational principle is adopted. The frequency-focusing method is applied to generate the extreme wave in the numerical wave tank. The wave load on the bottom mounted vertical cylinder is investigated. The hydroelastic response of the vertical cylinder is also investigated so as to compare the wave loads with the rigid body case in the extreme wave condition.

• The KIPS Transactions:PartA
• /
• v.10A no.3
• /
• pp.189-198
• /
• 2003

In this paper, we propose a method of analysis to show the heavy-tailed statistical distribution of file sizes in web servers, using TTT plot technique. TTT plot technique, a well-known method in the area of reliability engineering, determines that a distribution of samples fellows a heavy tailed one when their TTT statistical plots are lied on a straight line. We performed an intensive simulation using data gathered from real web servers. The simulation indicates that the proposed method is superior to Hill estimation technique or LLCD plot method in efficiency of data analysis. Moreover, the proposed method eliminates the possible decision error, which Pareto distribution or traditional method might cause.

• Environmental and Resource Economics Review
• /
• v.7 no.2
• /
• pp.31-51
• /
• 1998

우편설문조사를 이용한 조건부가치측정법을 통해 자료를 수집할 경우 무응답자의 발생은 보편적이며, 그러한 무응답자의 존재는 무응답자편의(無應答者偏倚) 가능성을 내재하고 있으므로, 표본의 통계치를 이용한 모집단에 대한 확대 적용을 위해서는 우선적으로 무응답자편의(無應答者偏倚)에 대한 검정(檢定)이 이루어져야 하며. 또한 무응답자편의(無應答者偏倚) 발생시 이에 대한 적절한 조치가 취해져야 한다. 본 연구에서는 다단계 우편설문발송법을 통해 수집된 자료를 이용하여 효과적인 무응답자편의(無應答者偏倚) 검정(檢定)의 수행과 무응답자군(無應答者群)의 평균지불의사액을 별도로 추정한 후 이를 이용하여 보다 합리적이고 정확한 총지불의사액(Aggregate WTP)의 도출방법을 제시하였다. 이는 우선적으로 무응답자편의(無應答者偏倚)를 분산분석을 통해 검정(檢定)한 후 선형외삽법으로 무응답자군(無應答者群)의 평균지불의사액을 추정하여 이를 모집단의 총지불의사액을 산출하는데 이용하는 방법으로 기존의 보수적인 방법들보다 상대적으로 저렴한 비용으로 무응답자편의(無應答者偏倚) 검정(檢定)을 가능하게 하며, 사회적 또는 인구특성상의 차이를 이용한 통계적 가중치 이용방법과는 달리 연구자의 직접적인 관심의 대상이 되는 지불의사액을 이용하여 무응답자편의(無應答者偏倚) 테스트를 하고 또한 무응답자군(無應答者群)의 평균지불의사액을 추정하여 보다 효율적이며 타당성이 있는 총지불의사액을 산출해 낼 수 있다는 장점이 있다.

• The KIPS Transactions:PartD
• /
• v.12D no.1 s.97
• /
• pp.103-110
• /
• 2005

Software Development is converting from structural to object oriented method. The later software development prefers the iterative process applications, not aterfall process and based on use case model, the requirements are expressed and based on this, analysis, design and coding are accomplished. Therefore, size of the software to be developed is estimated basing on use case and it is only possible to maintain the project success by estimating development effort, cost and development period. Even though development effort estimation models related current use case point. there is no appropriate development effort estimating. This paper shows, as a result of applying the development effort estimating model about UCP to the growth curve, a superior performance improvement to current statistical models. Therefore, estimation of development effort by applying this model, project development maintenance can be appropriately carried out.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2012.05a
• /
• pp.415-415
• /
• 2012

유역의 홍수위험도를 평가하기 위하여 선행 연구로서 DPSIR framework를 적용한 인자를 선정하였다. 이때 인자의 범위를 사회적, 경제적, 환경적 분야로 구분하고 각 분야에 대하여 홍수로 인하여 야기되는 유역의 다양한 영향요소를 포함하여 인자를 선정하였다. 본 연구에서는 이를 대상으로 각 인자의 가중치를 선정하기 위하여 패널식 조사연구기법인 Delphi 방법을 적용하였다. Delphi 기법은 델파이 절차가 반복되는 동안 조사 참여자에게 전회의 조사 결과, 즉 통계적 집단 반응을 피드백하며 참여자는 이를 참고하여 자신의 판단을 수정 보완 할 수 있는 방법이다. 따라서 1차 설문조사를 통하여 각 패널은 홍수위험도 평가 인자에 대해서 주관적인 가중치를 선정하고 이를 분석한 결과를 피드백 받아 참고한 후 2차 설문조사에 응하여 가중치를 재선정하게 된다. 조사에 참여할 패널의 선정은 수자원분야의 전문가를 대상으로 하였다. 또한 각 인자에 대한 가중치를 수집하는 방법을 다양하게 설정하여 기법에 따른 영향을 검토하였다. 수집기법은 상대적 가중치 결정법 중에서 가장 많이 사용되는 순위법과 직접입력법, 절대적인 기준에 대한 가중치를 수집하기 위하여 Fuzzy 가중치 입력법을 사용하였으며 이 때 각 인자가 소속되는 분야와 DPSIR framework 요소에 대해서도 별도로 가중치를 수집하여 최종적인 인자의 가중치를 결정하였다. 본 연구는 추가 연구를 통해 남한강에 적용되며 인자의 자료는 문헌조사, 통계자료 조사 및 수리학적 수치모형 등의 분석을 통해 수집될 예정이다. 이를 통하여 사회, 경제, 문화, 환경적인 측면을 고려한 인자들을 반영한 하천의 구간별 상대적 홍수취약도를 정량적으로 제시할 수 있으며 하천 구간별 치수관리 우선순위를 결정하는데 적극적으로 활용될 수 있다.

• Journal of Korean Society of Transportation
• /
• v.23 no.5 s.83
• /
• pp.15-25
• /
• 2005

Using a stochastic cost frontier approach(SCFA), this paper annually estimates the efficiency and productivity with same data in the papers which analyze the efficiency and productivity using data envelopment analysis(DEA) to compare the results and suggest the political findings of raising the efficiency and productivity for three publicly-owned rail transit properties. the Seoul Subway Corporation (SSC), the Seoul Metropolitan Electrified Railways Sector of Korea National Railroad (SMESRS) and the Busan Urban Transit Authority (BUTA). The results show that the results of SCFA are higher than DEA for efficiency and lower for productivity in that DEA regards the stochastic error and measurement error as the inefficiency contrary to SFCA. But the political findings from these results appears to be similar as follows. First, the productivity of the three properties should be first improved by using existing technologies efficiently and then by introducing new ones. Second, the three properties should improve the technical efficiency through reducing input quantities to raise their efficiency. Finally, all the three components of the productivity such as productive efficiency change, technical change, and scale change should be considered to evaluate their productivity more correctly.

• Journal of Advanced Navigation Technology
• /
• v.4 no.2
• /
• pp.162-170
• /
• 2000

Maneuvering targets are difficult for the Kalman filter to track since the target model of tracking filter might not fit the real target trajectory and the statistical characteristics of the target maneuver are unknown in advance. In order to track such a wildly maneuvering target, several schemes had been proposed and improved the tracking performance in some extent. In this paper a Kalman filter-based scheme is proposed for maneuvering target tracking. The proposed scheme estimates the target acceleration input vector directly from the feature of maneuvering target trajectories and updates the simple Kalman tracker by use of the acceleration estimates. Simulation results for various target profiles are analyzed for a comparison of the performances of our proposed scheme with that of conventional trackers.