http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
A Buckley - James Type Estimator of Quantile Function
송재기(Jae Kee Song),정동명(Dong Myeong Jeong) 한국데이터정보과학회 1992 한국데이터정보과학회지 Vol.3 No.2
We propose an estimator of quantile function by using the idea of Buckley-James(1979) that is used in estimating parameters in the linear regression with censored dependent variable. The Monte Carlo Simulation is performed to compare the proposed estimator with the Kaplan-Meier type estimator and the Nelson-Aalen type estimator, and it is shown that the proposed estimator is slightly better that other estimators for large y in most cases.
깁스 표본 기법을 이용한 베이지안 계층적 모형: 야생쥐의 예
송재기,이군희,하일도,Song, Jae-Kee,Lee, Gun-Hee,Ha, Il-Do 한국데이터정보과학회 1996 한국데이터정보과학회지 Vol.7 No.2
In this paper, we applied bayesian hierarchical model to analyze the field mice example introduced by Demster et al.(1981). For this example, we use Gibbs sampler method to provide the posterior mean and compared it with LSE(Least Square Estimator) and MLR(Maximum Likelihood estimator with Random effect) via the EM algorithm.
On The Number of Replications in Simulation Study
송재기,Song, Jae-Kee Korean Data and Information Science Society 1990 한국데이터정보과학회지 Vol.1 No.-
A method which determines the number of replications in the simulation is proposed, particularly for small-sample comparison of estimators. This method takes the smallest number of replications that makes the difference of mean square errors be statistically significant and provides an efficient algorithm for calculating the standard error of the mean square error. Two examples are illustrated, the first one is on comparison of mean and median ; the second, the Kaplan-Meier type and Buckley-James type estimators of a quantile function with censored data.
조현주,송명언,정동명,송재기,Cho, Hyun-Joo,Song, Myung-Unn,Jeong, Dong-Myung,Song, Jae-Kee 한국데이터정보과학회 1996 한국데이터정보과학회지 Vol.7 No.2
In this paper, we studied about several methods of testing hypothesis of correlation, specially Approximate method, Empirical method and Bootstrap method. The Approximate method is based on the Fisher's Z-transformation and the Empirical and Bootstrap methods approximate the distribution of the sample correlation coefficient by Monte Carlo simulation and Bootstrap technique, respectively. In order to compare how good these tests are, we computed powers under various alternatives. Consequently, we see that the Approximate test performs very well even if in small sample and all tests have almost the same power in large sample.
이혜선,송재기,박해상,전치혁,Lee Hye-Seon,Song Jae-Kee,Park Hae-Sang,Jun Chi-Hyuck 한국통계학회 2006 응용통계연구 Vol.19 No.1
Independent component analysis (ICA) is a statistical method for transforming an observed high-dimensional multivariate data into statistically independent components. ICA has been applied increasingly in wide fields of spectrum application since ICA is able to extract unknown components of a mixture from spectra. We focus on application of ICA for separating independent sources and predicting each composition using extracted components. The theory of ICA is introduced and an application to a metal surface spectra data will be described, where subsequent analysis using non-negative least square method is performed to predict composition ratio of each sample. Furthermore, some simulation experiments are performed to demonstrate the performance of the proposed approach. 독립성분분석은 차원이 높은 다변량데이타로부터 기저구조를 형성하는 독립성분을 분리하는데 사용되는 기법으로서 패턴인식, 예측 등 2차적 분석을 위한 1차 분석단계에서 사용할 수 있다. 본 연구에서는 독립성분분석을 이용하여 여러 혼합물 데이터로부터 독립성분을 분리한 다음 각 구성성분의 혼합비율을 예측하는 절차를 제안한다. 적용예로서 도금강판의 엑스선 회절강도값으로부터 여러가지 상을 분리한 다음 비음최소자승법을 이용하여 각 상의 분율을 예측하였으며, 이러한 제안방안의 타당성 평가를 위하여 모의 실험을 실시하였다.
이원기,송명언,송재기,이성국,박성화,Lee, Won-Kee,Song, Myung-Unn,Song, Jae-Kee,Lee, Sung-Kook,Park, Sung-Hwa 한국데이터정보과학회 1999 한국데이터정보과학회지 Vol.10 No.1
많은 역학조사에서 두발을 이용하여 체내 중금속 오염 정도를 알아보는 방법이 유용한 것으로 보고 있다. 본 논문에서는 61구의 부검체를 이용하여 다변량기법으로 두발과 장기의 수은 및 카드늄 축적정도를 측정하여 상호간 관련성을 밝히고자 하였다. 그 결과 수은의 경우 두발에서의 축적정도와 2차 축적장기와는 통계적으로 유의한 상관성이 있는 것으로 나타났다. It is well known to use scalp hair as a signal of strengths of heavy metals contamination. In this paper, using the multivariate methods we examine the relationship between strengths of cadmium and mercury in scalp hair and cerebellum, cerebrum, heart, kidney, liver, lung and spleen from 61 korean autopsy subjects. As a result, there is statistically singnificant relationship between strengths of mercury contamination in scalp hair and secondary contaminated organs.
Comparisons on Approximating Methods for Distribution of Sample Variance
최영애,송재기,Choi, Young-Ae,Song, Jae-Kee Korean Data and Information Science Society 1990 한국데이터정보과학회지 Vol.1 No.-
The Edgeworth expansion, the Roy-Tiku method and the bootstrap method for approximating the distribution of the sample variance are compared through the Monte Carlo simulation study.
NTrend 1.0에 의한 낙동강 수질 장기변동 추세분석
유재정 ( Jae Jeong Yu ),신석호 ( Suk Ho Shin ),윤영삼 ( Young Sam Yoon ),송재기 ( Jae Kee Song ) 한국물환경학회 2010 한국물환경학회지 Vol.26 No.6
The effect of seasonality on water quality variation is very significant. Generally, it reduce the power of the trend extraction. A parametric time-series model was used for detecting trends in historic constituent concentration data. The effect of seasonality is able to remove from time series decomposition technique. According to such statistic methode, long-term water quality trend analysis system (NTrend 1.0) was developed by Nakdong River Water Environmental Research Center. The trend analysis of BOD variation was conducted with NTrend 1.0 at Goreong and Moolkum site in Nakdong river to show the effect of water quality management action plan. Power test of trend extraction was tried each case of `deseasonalized and deannulized` data and `deseasonalized` data. Analysis period was from 1989 to 2006, and it`s period was divided again three times, 1989~1993, 1994~1999 and 2000~2006 according to action plan period. The BOD trend was downward in Goreong site during three times and it`s trend slope was very steep, and upward in Moolkum during 1989~1993, but it was turned downward during 1994~1999 and 2000~2006. It was revealed that it`s very effective to reduce the concentration of BOD by water quality management action plan in that watershed. The result of power test was shown that it is high for trend extraction power in case of `deseasonalized` data.
이원기,송명언,송재기,박희주,Lee, Won-Kee,Song, Myung-Unn,Song, Jae-Kee,Park, Hee-Joo 한국데이터정보과학회 1998 한국데이터정보과학회지 Vol.9 No.1
임의중도절단된 생존시간자료에서 생존함수에 대한 동시신뢰대를 근사식이나 표없이 구성하는 간단한 방법을 제안하였다. 그리고 모의실험을 통하여 기존의 동시신뢰대와 포함확률측면에서 서로 비교하고, 실제자료에 적용하여 보았다. We consider the problem of obtaining the confidence bands for the survival function with incomplete data. It is a rather simple procedure for constructing confidence bands of survival function. This method uses the weak convergence of normalized cumulative hazard estimator to a mean zero Gaussian process whose distribution can be easily approximated through simulation. Finally, we compare the performance of the proposed confidence bands through Monte Carlo simulation and we applied to construct the proposed bands with the Leukemia patient data.