• The Korean Journal of Financial Studies
• /
• v.11 no.1
• /
• pp.157-176
• /
• 2005

본 연구는 KOSPI자산 포트폴리오에 대한 VaR를 다양한 ARCH류 모형을 사용하여 추정하고 이들의 예측능력을 평가하였다. 활용된 모형은 우선 기본적인 GARCH(1,1)모형과 레버리지 효과를 감안한 TGARCH모형, 다양한 ARCH모형을 포괄할 수 있는 PGARCH모형, 변동성의 영속성을 고려한 IGARCH모형이 포함되었다. 모형 상호간의 성과비교에 추가하여 ARCH류 모형에서 수익률예측오차의 분포에 따라서 VaR의 예측성과가 얼마나 차이가 발생하는가를 확인하기 위하여 정규분포와 Student-t분포의 성과를 비교하였다. 마지막으로 VaR 추정시에 조건부평균을 무시하는 관례가 어느정도 타당성이 있는지를 확인하기 위하여 1시차 자기회귀과정에 입각한 조건부 평균을 감안한 결과를 검토하였다. ARCH류 모형에서 모형 설명력은 보다 정교한 모형인 TGARCH모형이나 PGARCH모형이 우월하게 나타났지만, VaR의 예측능력 우월성으로 이어지지는 않았다. Student-t분포를 가정한 경우 VaR모형 사후검증성과는 정규분포를 가정한 경우보다 모든 신뢰수준에서 개선되었으며, 조건부평균의 제거는 Student-t분포 가정하에서는 적합하지 않은 것으로 나타났다. ARCH류 모형에서 가장 단순한 형태인 IGARCH모형의 예측성과가 다른 모형들에 비하여 뒤떨어지지 않으며, 더욱 제약된 형태인 RiskMetrics의 EWMA모형이 사후검증에서 우수한 성과를 보여 단순한 모형의 유용성을 확인시켜주고 있다.

• International Journal of Highway Engineering
• /
• v.9 no.3
• /
• pp.13-20
• /
• 2007

For calculating the number of lane, it is essential to gain the 30th or 100th highest design hourly volume. The design hourly volume obtained from AADT multiplied by design hour factor. In this paper, we developed the regression models fur estimating the 30th highest hour volume and 100th highest hour volume as defined by AADT 50,000 criterion based on the data obtained the 34 monitoring sites in highway. By comparing the performance of the proposed models and conventional models using MAPE, the proposed model for 30th highest design hourly volume reduced the estimator error of 11.83% than that of conventional methods for less than AADT 50,000 and decreased estimation error of 22.17% than that of conventional method for more than AADT 50,000. Moreover, the proposed model for 100th highest design hourly volume reduced the estimator error of 8.16% than that of conventional methods for less than AADT 50,000 and decreased estimation error of 15.25% than that of conventional method for more than AADT 50,000.

• Journal of Korean Society of Transportation
• /
• v.15 no.4
• /
• pp.53-71
• /
• 1997

This paper presents a methodology for estimating average travel speed using volume and occupancy data from single magnetic loop detectors for signalized arterials. Three methods were developed and evaluated using field data: VPLUSKO method, fuzzy control method, and neural network method. While the VPLUSKO method is easy to apply, it results poor performances compared to other methods. The neural network method showed the best performances among the candidate methods. This method revealed the weakness in transferability, however. From limited cases of field test, it was concluded that the method of the fuzzy control application showed reasonable performance of estimation. It was also demonstrated that the fuzzy control method has the capability of transferability.

• Journal of Animal Science and Technology
• /
• v.50 no.1
• /
• pp.19-26
• /
• 2008

Heritability and genetic correlation for growth traits in Duroc pig breed were estimated using Bayesian method via Gibbs sampling. The data set consisted of 3,526 performance records at National Institute of Animal Science. For estimating those parameters using Gibbs sampling, 5,000 cycles of ‘burn-in’ period were discarded among a total of 55,000 samples. Out of the remaining 50,000 samples, 5,000 estimates by each parameter were retained and used for analyses to avoid any correlation among adjacent samples. The growth traits considered in this study were average daily gain at 30kg(ADG1), average daily gain at 90kg(ADG2), backfat thickness(BF), days to 90kg(D90) and feed conversion ratio(FC). The estimated heritabilities and their standard deviation using Gibbs sampler were 0.43±0.04, 0.49±0.038, 0.31±0.040, 0.48±0.039 and 0.62±0.086, respectively. Genetic correlations were －0.02, －0.13, －0.55 and －0.15 between ADG1 with ADG2, BF, D90 and FC, respectively, 0.16, －0.73, －0.32 between ADG2 with BF, D90 and FC respectively, 0.01, －0.08 between BF with D90, FC, respectively, and 0.23 between D90 with FC.

• Proceedings of the KSEEG Conference
• /
• 2007.04a
• /
• pp.537-537
• /
• 2007

• Proceedings of the Korea Society of Poultry Science Conference
• /
• 2002.11a
• /
• pp.140-142
• /
• 2002

This study was conducted to obtain some genetic information for developing superior Korean native chickens. Heterosis and combining ability effects were estimated with 1,080 progenies of full diallel crosses of 3 strains in Korean Native Chicken. The traits of egg productivity were Investigated, and heterosis effect, general combining ability, specific combining ability and reciprocal effect were analyzed model I of the Griffing's method 1. The general performance were 147.0 days in age at first egg laying, 71.8 eggs in egg numbers to 270 days, 50.1g in egg weight at 270 days, 75.6 in egg shape index, 356$\mu\textrm{m}$ in egg shell thickness, and 4.27kg/$\textrm{cm}^2$ in egg shell strength, respectively. According to result, heterosis effect revealed comparatively high in age at first egg laying, egg numbers, egg weight, and egg shell strength. The effects estimated highly in general combining ability for age at first egg laying, egg numbers, egg weight, egg shape Index and egg shell thickness, and in specific combining ability were egg numbers, egg weight and egg shell thickness, and in reciprocal effects for egg numbers, egg shape index, and egg shell thickness. This result could be used for developing Korean native chicken.

• The Mathematical Education
• /
• v.33 no.2
• /
• pp.167-175
• /
• 1994

결측치를 가진 회귀모형의 모수 추정법을 이용하여 목표지향형 평가 모델에서 기초고사(X)와 신고사(I)(Y), 신고사(II)(Z)등 두 개 이상의 고사지로 이루어진 고사집에서 기초고사에는 결측치가 없고 신고사(I), 신고사(II)등에는 결측치가 있는 경우 모수의 최우추정량 계산법을 논하고 E.M. 알고리즘과 평가치는 희귀방적식화에 의하여 우리나라 중등학교 학생의 수학학습능력과 수학적 사고력의 크기를 변별하며 학생들의 진능력이 반영된 평가모델과 최종 성적을 평가 할 수 있는 계산법을 제시하였다.

• Proceedings of the Korea Society for Simulation Conference
• /
• 1999.04a
• /
• pp.32-36
• /
• 1999

본 연구에서는 대공 무기체계의 성능평가를 위한 시뮬레이션 모델의 개발로서 능력추정을 위한 수식의 전개와 전산프로그램을 개발하고 예제를 통하여 능력산출 예를 보였다. 본 연구에서 고려된 시뮬레이션 방법으로 Mon Carlo 시뮬레이션과 Karman Filtering을 사용하였다. 본 모델의 주요 단계로서 1) 가장 최단 경로에서의 표적과 탄두간의 상대 속도와 위치를 결정하고, 2) 표적의 탄두에 대한 취약면적을 산출하고 3) 마지막으로 표적살상 확률을 산정하는 과정의 단계에 따라 개발되었다.

• KDI Journal of Economic Policy
• /
• v.32 no.2
• /
• pp.27-53
• /
• 2010

This paper examines whether and how much vocational training raises an individual's earnings in Korea, using the Economically Active Population Survey. To overcome endogeneity of training, we apply fixed-effects and propensity-score matching (PSM) methods. Fixed-effects (PSM) results suggest that work-related training received in the previous one year increases a worker's monthly earnings by 2.6 to 4.7 (7.5 to 9.8) percent. Taken altogether, work-related training enhances a worker's earnings by a minimum of 2.6 and a maximum of 9.8 percent in Korea.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.27 no.1A
• /
• pp.51-57
• /
• 2002

In this paper, a simple interpolation technique in a frequency domain is proposed for the discrete Fourier transform(DFT) algorithm. Frequency and phase resolution capabilities of the DFT algorithm can be significantly improved by the proposed interpolation technique without increase of a DFT size(the number of points for the DFT). The new technique uses a diving point in amplitude and phase spectrums. As an application, the technique can be used for joint estimation of fine frequency and phase offsets in burst mode digital transmission. Simulation results show that the joint estimator using the technique is robust to estimation errors.