• 응용통계연구
• /
• 제27권6호
• /
• pp.1039-1047
• /
• 2014

시계열 자료를 위한 가장 기본적인 모형인 자기회귀모형을 고려한다. 흔히 시계열 자료에서 정규성 가정이 위배되는 경우가 발생하며, 정규성 가정을 완화하기 위한 방법으로 두꺼운 꼬리를 가지는 분포 또는 비대칭 분포를 고려할 수 있다. 비대칭 지수멱 분포의 사용은 비뚤림이 있는 두꺼운 꼬리를 가지는 자기회귀모형의 이상치의 영향을 줄이고 로버스트한 추론을 할 수 있도록 한다. 본 논문에서는 자기회귀모형에 대한 오차항에 정규분포 보다 첨도와 왜도에 유연함을 가지는 분포를 고려함으로써 정규성 가정을 완화하여 추론하고자 하였다. 정규분포의 대안으로 비대칭 지수멱 분포를 고려하였으며 정규분포의 결과와 비교 하여 비대칭 지수멱 분포의 로버스트함을 보였다. 또한 주어진 분포에 대한 효율적인 베이지안 추론을 하기 위하여 SIR 알고리즘과 격자망 방법을 고려하였다.

• Communications for Statistical Applications and Methods
• /
• 제24권2호
• /
• pp.129-142
• /
• 2017

In this paper we consider the problem of the model selection/discrimination among three different positively skewed lifetime distributions. Lindley, Weibull, and gamma distributions have been used to effectively analyze positively skewed lifetime data. This paper assesses how much closer the Lindley distribution gets to Weibull and gamma distributions. We consider three techniques that involve the likelihood ratio test, asymptotic likelihood ratio test, and minimum Kolmogorov distance as optimality criteria to diagnose the appropriate fitting model among the three distributions for a given data set. Monte Carlo simulation study is performed for computing the probability of correct selection based on the considered optimality criteria among these families of distributions for various choices of sample sizes and shape parameters. It is observed that overall, the Lindley distribution is closer to Weibull distribution in the sense of likelihood ratio and Kolmogorov criteria. A real data set is presented and analyzed for illustrative purposes.

• 한국물환경학회지
• /
• 제39권4호
• /
• pp.302-315
• /
• 2023

Distribution characteristics and altitudinal gradient of four species (E. strigata, E. separigata, E. orientalis-sachalinensis group) of the mayfly genus Ephemera (Order Ephemeroptera) were analyzed with probability distribution models (exponential, normal, lognormal, logistic, Weibull, gamma, beta, Gumbel). Data was collected from 23,846 sampling units of 6,787 sites in Korea from 2010 to 2021. The beta distribution model showed the best fit for positively skewed E. orientalis-sachalinensis and little-skewed E. strigata along with altitudinal gradient. The reversed lognormal distribution model showed the best-fit for negatively skewed E. separigata. E. orientalis-sachalinensis distributed at the range of altitude 1~700 m (mean 251 m, median 226 m, mode 124 m, and standard deviation 161 m), E. strigata distributed at the range of altitude 5~871 m (mean 474 m, median 478 m, mode 492 m, and standard deviation 200 m), E. separigata distributed at the range of altitude 7~846 m (mean 620 m, median 659 m, mode 760 m, and standard deviation 181 m). Altitudinal habitat suitability ranges were estimated to be 42~257 m for E. orientalis-sachalinensis, 335~644 m for E. strigata, and 641~824 m for E. separigata. Based on the altitudinal spectrum of suitability and altitude-related temperature analysis results, E. orientalis-sachalinensis was estimated to be thermophilic, E. strigata to be mesophilic, and E. separigata to be thermophobic. This is the first national-scale evaluation of the altitudinal distribution of Ephemera in Korea. These results will be used in a further research study on altitudinal shift of the species of Ephemera under climate change.

• Communications for Statistical Applications and Methods
• /
• 제30권3호
• /
• pp.227-244
• /
• 2023

In this paper, a depth-based nonparametric test for a multivariate two-sample scale problem is proposed. The proposed test statistic is based on the depth-induced ranks and is thus distribution-free. In this article, the depth values of data points of one sample are calculated with respect to the other sample or distribution and vice versa. A comprehensive simulation study is used to examine the performance of the proposed test for symmetric as well as skewed distributions. Comparison of the proposed test with the existing depth-based nonparametric tests is accomplished through empirical powers over different depth functions. The simulation study admits that the proposed test outperforms existing nonparametric depth-based tests for symmetric and skewed distributions. Finally, an actual life data set is used to demonstrate the applicability of the proposed test.

• 응용통계연구
• /
• 제16권1호
• /
• pp.101-115
• /
• 2003

본 논문에서는 Chen, Dey와 Shao(1999), Branco와 Dey(2001)가 제안한 왜도가 있는 두터운 꼬리를 가지는 오차 분포와 디리슈레 과정 사전분포를 이용한 베이지안 메타분석 (meta-analysis)을 하고자 한다. 베이지안 메타분석을 위하여 가중함수를 고려한 계층적 선택 모형을 이용한다. 이때의 오차항은 왜도가 있는 비정규 분포로 가정한다. 이를 위하여 우선 왜도 타원형 분포의 일반적인 족을 소개한다 이 분포족중 왜도 정규분포와 왜도 t 분포를 오차항 분포로 이용한 베이지안 계층적 선택 모형을 고려하며, 이 때 발생하는 복잡한 베이지안 계산은 MCMC 방법으로 해결한다. 마지막으로, 실제 자료(Johnson, 1993)인 두 가지의 충치예방약의 효과에 대한 차이를 비교하기 위해 얻어진 12개의 연구 자료를 이용하여 본 연구에서 제시된 베이지안 방법을 이용하여 메타분석을 한다.

• 소음진동
• /
• 제8권5호
• /
• pp.870-878
• /
• 1998

The cross-flow fan which constitutes a fan-duct system with a stabilizer and a scroll casing is widely used in many air-ventilating and air-conditioning devices. Its ooperating points of high efficiency and loading conditions frequently induce a annoying sharp tonal component of discrete frequency on the noise spectrum, which is open called as a BPF(Blade-Passing-Frequency) noise and degrades the sound quality of the devices. this BPF tone has been one of the defects of the cross-flow fan. This study proposes two methods in order to reduce this tonal noise component, which are the random distributions of the fan blades and the skewed shapes of the stabilizer. The proposed methods are verified by a simple analytical model and are applied in manufacturing the cross-flow fan and the stabilizer samples. Some experiments are carried out to verify the reduction capability of BPF tones of above two schemes and the experimental results are analyzed. The comparison between two method is also carried out.

• 한국대기환경학회지
• /
• 제13권6호
• /
• pp.475-486
• /
• 1997

The characteristics of atmospheric aerosols were investigated as a function of particle size and water solubility. The atmospheric aerosols were sampled with classifying into 12 size ranges by the use of Andersen low-pressure impactor. Collected aerosol particles were extracted by ultrapure water and filtered to be separated into water-soluble and insoluble components. The concentrations 12 elements in both components were determined by PIXE analysis. And the concentrations of 8 ions in the soluble component were analyzed by ion chromatography. In general, the mass size distribution of particulate matter was represented as a bimodal distribution. The mass size distributions of S$(SO_4^{2-}), K(K^+), Zn and NH_4^+$ skewed to the smaller size range and those of Si, Ca$(Ca^{2+}), Fe, Na^+ and Mg^{2+}$ skewed to the larger size range. They had roughly one peak in the fine and coarse particle region,respectively. On the other hand, the mass size distribution of Ti, Mn, Ni, Cu, $Cl^- and NO_3^-$ were represented as the bimodal distribution. Fe and Si in the aerosol particles extracted into pure water are existing in high insoluble state. Conversely, almost the whole of S is dissolved in water.

• 한국경영과학회:학술대회논문집
• /
• 대한산업공학회/한국경영과학회 2000년도 춘계공동학술대회 논문집
• /
• pp.365-368
• /
• 2000

This paper proposes a new heuristic method of constructing process capability indices (PCIs) for skewed populations. It is based on weighted standard deviation (WSD) method which decomposes the standard deviation of a quality characteristic into upper and lower deviations and adjusts the value of PCI using decomposed deviations in accordance with the skewness estimated from sample data. For symmetric populations, the proposed PCIs reduce to standard PCIs. Asymptotic distributions of the estimators of the PCIs are obtained. The performances of the proposed methods are compared with those of the standard and other methods. Numerical comparisons indicate that considerable improvements over existing methods can be achieved by the use of WSD method when the underlying distribution is skewed.

• 한국농공학회지
• /
• 제34권4호
• /
• pp.39-47
• /
• 1992

Derivation of reasonable design low flows was attempted by comparative analysis of design low flows was derived by Power and SMEMAX transformations for the normalizations of skewed distribution and by Type m extremal distribution presented in the first report of this study with annual low flows in the five watersheds of main river basins in Korea. The results were anslyzed and summarized as follows. 1.Basic statistics of annual low flows for the selected watersheds were calculated by using Power and SMEMAX transformations. 2.Power thansformation has found to be the best for the normalization of skewed distribution among others including log, square root and SMEMAX transformations. 3.Design low flows for the selected watersheds were derived by the Power and SMEMAX transformations. 4.Judging by the relative suitabilities of the Type III extremal distribution, Power and SMEMAX transformation, it was found that design low flows of all methods are closer to the observed data within 10 years of the return period and those of Power transformation can be acknowledzed as a reasonable one among others from the viewpoint of the median between values of Type m extremal distribution and SMEMAX transformation in addition to closing the observed than others over 10 years of the return period.

• 산업공학
• /
• 제17권1호
• /
• pp.1-12
• /
• 2004

In a modern semiconductor device manufacturing industry, statistical bin limits on wafer level test bin data are used for minimizing value added to defective product as well as protecting end customers from potential quality and reliability excursion. Most wafer level test bin data show skewed distributions. By Monte Carlo simulation, this paper evaluates methods and sample size effect regarding determination of statistical bin limits. In the simulation, it is assumed that wafer level test bin data follow the Poisson distribution. Hence, typical shapes of the data distribution can be specified in terms of the distribution's parameter. This study examines three different methods; 1) percentile based methodology; 2) data transformation; and 3) Poisson model fitting. The mean square error is adopted as a performance measure for each simulation scenario. Then, a case study is presented. Results show that the percentile and transformation based methods give more stable statistical bin limits associated with the real dataset. However, with highly skewed distributions, the transformation based method should be used with caution in determining statistical bin limits. When the data are well fitted to a certain probability distribution, the model fitting approach can be used in the determination. As for the sample size effect, the mean square error seems to reduce exponentially according to the sample size.