• 천문학회보
• /
• 제35권2호
• /
• pp.51.2-51.2
• /
• 2010

We have investigated 63 intense geomagnetic storms (Dst $\leq$ -100 nT) that occurred from 1998 to 2006. Using these events, we compared Dst forecast models: Burton et al. (1975), Fenrich and Luhmann (1998), O'Brien and McPherron (2000a), Wang et al. (2003), and Temerin and Li (2002, 2006) models. For comparison, we examined a linear correlation coefficient, RMS error, the difference of Dst minimum value (${\Delta}$peak), and the difference of Dst minimum time (${\Delta}$peak_time) between the observed and the predicted during geomagnetic storm period. As a result, we found that Temerin and Li model is mostly much better than other models. The model produces a linear correlation coefficient of 0.94, a RMS (Root Mean Square) error of 14.89 nT, a MAD (Mean Absolute Deviation) of ${\Delta}$peak of 12.54 nT, and a MAD of ${\Delta}$peak_time of 1.44 hour. Also, we classified storm events as five groups according to their interplanetary origin structures: 17 sMC events (IP shock and MC), 18 SH events (sheath field), 10 SH+MC events (Sheath field and MC), 8 CIR events, and 10 nonMC events (non-MC type ICME). We found that Temerin and Li model is also best for all structures. The RMS error and MAD of ${\Delta}$peak of their model depend on their associated interplanetary structures like; 19.1 nT and 16.7 nT for sMC, 12.5 nT and 7.8 nT for SH, 17.6 nT and 15.8 nT for SH+MC, 11.8 nT and 8.6 nT for CIR, and 11.9 nT and 10.5 nT for nonMC. One interesting thing is that MC-associated storms produce larger errors than the other-associated ones. Especially, the values of RMS error and MAD of ${\Delta}$peak of SH structure of Temerin and Li model are very lower than those of other models.

• Journal of the Korean Data and Information Science Society
• /
• 제15권4호
• /
• pp.743-752
• /
• 2004

In this paper, we develop the Bayesian model selection procedure using the reference prior for comparing two nested model such as the independent and intraclass models using the distance or divergence between the two as the basis of comparison. A suitable criterion for this is the power divergence measure as introduced by Cressie and Read(1984). Such a measure includes the Kullback -Liebler divergence measures and the Hellinger divergence measure as special cases. For this problem, the power divergence measure turns out to be a function solely of $\rho$, the intraclass correlation coefficient. Also, this function is convex, and the minimum is attained at $\rho=0$. We use reference prior for $\rho$. Due to the duality between hypothesis tests and set estimation, the hypothesis testing problem can also be solved by solving a corresponding set estimation problem. The present paper develops Bayesian method based on the Kullback-Liebler and Hellinger divergence measures, rejecting $H_0:\rho=0$ when the specified divergence measure exceeds some number d. This number d is so chosen that the resulting credible interval for the divergence measure has specified coverage probability $1-{\alpha}$. The length of such an interval is compared with the equal two-tailed credible interval and the HPD credible interval for $\rho$ with the same coverage probability which can also be inverted into acceptance regions of $H_0:\rho=0$. Example is considered where the HPD interval based on the one-at- a-time reference prior turns out to be the shortest credible interval having the same coverage probability.

• 정보과학회 컴퓨팅의 실제 논문지
• /
• 제22권9호
• /
• pp.461-466
• /
• 2016

토픽 확장은 학습된 토픽의 질을 향상시키기 위해 추가적인 외부 데이터를 반영하여 점진적으로 토픽을 확장하는 방법이다. 기존의 온라인 학습 토픽 모델에서는 외부 데이터를 확장에 사용될 경우, 새로운 단어가 기존의 학습된 모델에 반영되지 않는다는 문제가 있었다. 본 논문에서는 무한 사전 온라인 LDA 토픽 모델을 이용하여 외부 데이터를 반영한 토픽 모델 확장 방법을 연구하였다. 토픽 확장 학습에서는 기존에 형성된 토픽과 추가된 외부 데이터의 단어와 유사도를 반영하여 토픽을 확장한다. 실험에서는 기존의 토픽 확장 모델들과 비교하였다. 비교 결과, 제안한 방법에서 외부 연관 문서 단어를 토픽 모델에 반영하기 때문에 대본 토픽이 다루지 못한 정보들을 토픽에 포함할 수 있었다. 또한, 일관성 평가에서도 비교 모델보다 뛰어난 성능을 나타냈다.

• Geomechanics and Engineering
• /
• 제23권1호
• /
• pp.51-59
• /
• 2020

In this paper, the WEKA platform was used to mine and analyze measured data of floor failure depth and a prediction system of floor failure depth was developed with Java. Based on the standardization and discretization of 35-set measured data of floor failure depth in China, the grey correlation degree analysis on five factors affecting the floor failure depth was carried out. The correlation order from big to small is: mining depth, working face length, floor failure resistance, mining thickness, dip angle of coal seams. Naive Bayes model, neural network model and decision tree model were used for learning and training, and the accuracy of the confusion matrix, detailed accuracy and node error rate were analyzed. Finally, artificial neural network was concluded to be the optimal model. Based on Java language, a prediction system of floor failure depth was developed. With the easy operation in the system, the prediction from measured data and error analyses were performed for nine sets of data. The results show that the WEKA prediction formula has the smallest relative error and the best prediction effect. Besides, the applicability of WEKA prediction formula was analyzed. The results show that WEKA prediction has a better applicability under the coal seam mining depth of 110 m~550 m, dip angle of coal seams of 0°~15° and working face length of 30 m~135 m.

• 천문학회지
• /
• 제37권4호
• /
• pp.151-157
• /
• 2004

It is investigated quantitative relations between the magnetic storm magnitude and the solar wind parameters such as the Interplanetary Magnetic Field (hereinafter, IMF) magnitude (B), the southward component of IMF (Bz), and the dynamic pressure during the main phase of the magnetic storm with focus on the role of the interplanetary shock (hereinafter, IPS) in order to build the space weather fore-casting model in the future capable to predict the occurrence of the magnetic storm and its magnitude quantitatively. Total 113 moderate and intense magnetic storms and 189 forward IPSs are selected for four years from 1998 to 2001. The results agree with the general consensus that solar wind parameter, especially, Bz component in the shocked gas region plays the most important role in generating storms (Tsurutani and Gonzales, 1997). However, we found that the correlations between the solar wind parameters and the magnetic storm magnitude are higher in case the storm happens after the IPS passing than in case the storm occurs without any IPS influence. The correlation coefficients of B and $BZ_(min)$ are specially over 0.8 while the magnetic storms are driven by IPSs. Even though recently a Dst prediction model based on the real time solar wind data (Temerin and Li, 2002) is made, our correlation test results would be supplementary in estimating the prediction error of such kind of model and in improving the model by using the different fitting parameters in cases associated with IPS or not associated with IPS rather than single fitting parameter in the current model.

• 산업경영시스템학회지
• /
• 제40권3호
• /
• pp.92-98
• /
• 2017

The rapid growth of engineering technology and the emergence of systemized and large-scale engineering systems have resulted in complexity and uncertainty throughout the lifecycle activities of engineering systems. This complex and large-scale engineering system consists of numerous components, but system failure can be caused by failure of any one of a number of components. There is a real difficulty in managing such a complex and large-scale system as a part. In order to efficiently manage the system and have high reliability, it is necessary to structure a system with a complex structure as a sub-system. Also, in the case of a system in which cause of failures exist at the same time, it is required to identify the correlation of the components lifetime and utilize it for the design policy or maintenance activities of the system. Competitive risk theory has been used as a theory based on this concept. In this study, we apply the competitive risk theory to the models with combined structure of series and parallel which is the basic structure of most complex engineering systems. We construct a competing risks model and propose a mathematical model of net lifetime and crude lifetime for each cause of failure, assuming that the components consisting a parallel system are mutually dependent. In addition, based on the constructed model, the correlation of cause of failure is mathematically analyzed and the hazard function is derived by dividing into net lifetime and crude lifetime.

• 대한지리학회지
• /
• 제46권3호
• /
• pp.319-330
• /
• 2011

본 연구에서는 지상관측 토양수분, 강수량, 지면온도 및 MODIS NDVI와 인공신경망모형을 이용하여 토양수분 공간분포 산정 모형을 제안하였으며, 신뢰성 높은 토양수분 관측 자료를 보유한 용담댐 유역에 대하여 모형의 적용성을 검증하였다. 토양수분 산정모형의 학습에 사용된 주천, 부귀, 상전의 3개 지점의 경우 약 0.9353의 상관계수와 약 1.4957%의 평균제곱근오차를 보여주며, 검증지점으로 사용된 천천2의 경우에는 약 0.8215의 상관계수와 약 4.2077%의 평균제곱근오차를 보여 토양수분 산정모형의 적용가능성이 높다고 판단된다. 인공위성으로부터 관측된 광역의 식생정보와 자료간의 비선형 상관특성을 잘 구현하는 인공신경망을 활용하여 수립된 토양수분 산정모형을 이용하여 용담댐 유역의 토양수분 공간분포도를 산정한 결과, 용담댐 유역의 대부분을 차지하고 있는 산림지역의 토양수분이 다른 지역에 비하여 높은 수치를 보여주는 토양수분의 분포를 보여주었다. 본 연구를 통해 제시된 토양수분 산정 방법은 광역 토양수분 산정에 유용한 접근법으로 판단된다.

• 재무관리연구
• /
• 제21권2호
• /
• pp.183-209
• /
• 2004

본 연구에서는 금리변동에 따른 헤지수단의 목적으로 도입된 국채선물과 해당 기초자산인 국채현물의 일별 자료를 통해 두 시계열의 상호관계를 변동성에 초점을 두고 Bivariate GARCH 모형인 BEKK 모형과 국면전환 및 백터 오차수정항이 포함된 Bivariate-AR(1)-Markov-Switching-VECM 모형을 이용하여 비교 분석하였다. 본 연구의 분석기간은 2000년 1월 4일부터 2003년 10월 30일까지이며 분석대상은 일별 국채현물지수와 국채선물지수 935 관측치 이다. 본 연구의 결과 우리나라에서 국채선물과 현물시장의 분석에 있어서 두 시장을 한꺼번에 아우를 수 있는 Bivariate 모형설정의 중요성이 강하게 대두되었다. 특히 본 연구의 분석기간 중에는 국채시장의 상승국면과 하락국면이라는 두 상태보다는 국채가격의 변동성국면이 훨씬 더 강하게 국채시장에 작용하고 있음이 밝혀졌다. 이는 투자자가 보다 나은 헷징결과를 기대한다면 국채시장의 분석시 현물과 선물, 각각의 분산과정뿐만 아니라 공분산과정도 반드시 시계열모형내에서 동시에 고려해야함을 시사하고 있다.

• 한국소음진동공학회논문집
• /
• 제22권6호
• /
• pp.509-523
• /
• 2012

The sound quality based on design optimization, throughout the development process of various electronic office equipments, needs to be considered in order to respond the increased needs for the emotional satisfaction of customers in terms of psycho-acoustics. This paper focuses on how to describe the characteristics of operating sound radiated from laser printers by using various sound attributes, and to model the sound quality index that can properly evaluate the subjective preference on modification conditions in the improvement study quantitatively. Especially, the proposed verification process, in the form of combining the correlation based method and the decision error based method, was applied to improve the generality and reliability of a group of participants in the jury evaluation. The modified Aures tonality model was also proposed to improve the correlation coefficient with the mean response of participants by optimizing some parameters. As a result, the loudness, articulation index, roughness, tonality, fluctuation strength were used to model the sound quality index for laser printers by using the multiple-linear regression method. Through the improvement study, it was confirmed that replacing the absorbing materials is effective to reduce the tonalness radiated from the side of a reference printer model. Based on above results, it can be concluded that the proposed model has enough usefulness as quantitative evaluation index to evaluate the difference between modification conditions in the improvement study.

• 인터넷정보학회논문지
• /
• 제14권2호
• /
• pp.25-34
• /
• 2013

본 논문에서는 워크플로우 기반 인적 자원의 소속성 분석을 위한 업무-수행자 이분 행렬 생성 알고리즘을 제안한다. 워크플로우 기반 인적 자원은 워크플로우 관리 시스템에 의해 관리되는 조직의 모든 수행자들을 말하며, 워크플로우 모델의 실행 과정에서 특정 업무 집합에 참여하게 된다. 이러한 워크플로우 모델에 정의된 수행자들과 업무들과의 소속성을 나타내는 소셜 네트워크를 업무-수행자 소속성 네트워크라 정의하였으며, 본 논문에서 제안하는 알고리즘은 워크플로우 모델로부터 발견된 업무-수행자 소속성 네트워크 모델(APANM)에 대한 이분 행렬을 생성하기 위한 알고리즘이다. 결론적으로, 알고리즘에 의해 생성된 업무-수행자 이분 행렬은 중심성(centrality), 밀집도(density), 상관 관계(correlation)와 같은 다양한 소셜 네트워크 관련 속성들을 분석하는데 적용될 수 있으며, 이를 통해 워크플로우 기반 인적 자원의 소속성에 대한 유용한 지식을 획득할 수 있다.