• The Korean Journal of Applied Statistics
• /
• v.18 no.2
• /
• pp.381-393
• /
• 2005

As a promising technique for dimension reduction in regression analysis, Sliced Inverse Regression (SIR) and an associated chi-square test for dimensionality were introduced by Li (1991). However, Li's test needs assumption of Normality for predictors and found to be heavily dependent on the number of slices. We will provide a unified asymptotic test for determining the dimensionality of the SIR model which is based on the probabilistic principal component analysis and free of normality assumption on predictors. Illustrative results with simulated and real examples will also be provided.

• The Korean Journal of Applied Statistics
• /
• v.18 no.1
• /
• pp.43-56
• /
• 2005

In this study, a robust estimation method for the first-order autocorrelation coefficient in the time series model following AR(l) process with additive outlier(AO) is investigated. We propose the L-type trimmed least squares estimation method using the preliminary estimator (PE) suggested by Rupport and Carroll (1980) in multiple regression model. In addition, using Mallows' weight function in order to down-weight the outlier of X-axis, the bounded-influence PE (BIPE) estimator is obtained and the mean squared error (MSE) performance of various estimators for autocorrelation coefficient are compared using Monte Carlo experiments. From the results of Monte-Carlo study, the efficiency of BIPE(LAD) estimator using the generalized-LAD to preliminary estimator performs well relative to other estimators.

• Communications for Statistical Applications and Methods
• /
• v.16 no.1
• /
• pp.103-113
• /
• 2009

Weighting is a common form of unit nonresponse adjustment in sample surveys where entire questionnaires are missing due to noncontact or refusal to participate. A common approach computes the response weight as the inverse of the response rate within adjustment cells based on covariate information. In this paper, we consider the efficiency and robustness of nonresponse weight adjustment bated on the response propensity and predictive mean. In the simulation study based on 2000 Fishry Census in Korea, the root mean squared errors for assessing the various ways of forming nonresponse adjustment cell s are investigated. The simulation result suggest that the most important feature of variables for inclusion in weighting adjustment is that they are predictive of survey outcomes. Though useful, prediction of the propensity to response is a secondary. Also the result suggest that adjustment cells based on joint classification by the response propensity and predictor of the outcomes is productive.

• Analytical Science and Technology
• /
• v.26 no.3
• /
• pp.165-173
• /
• 2013

To establishment of PT Program for Indoor air quality field that manufacture of confidential development PTMs (proficiency testing materials) and examined of proficiency testing evaluation included sampling process whether or not that is valid. Confirmation of homogeneity and stability of PTMs prepared. PTMs were confirmed to be homogeneous enough to be used as proficiency testing materials since withinbottle homogeneities of the RMs were lower than 0.3 times of targeted standard deviation of proficiency testing. The result of this study showed that the Robust RSD of proficiency testing for VOCs (volatile organic compounds) appeared 23~43% in concentration of 50~320 ${\mu}g/m^3$ for Method A(Distribution by adsorption in Tenax-tube of VOCs), but less 13~42% in concentration of 200~1200 ${\mu}g/m^3$, 16~31% in concentration of 100~450 ${\mu}g/m^3$ for Method B (distribution by VOCs of gas phase in 10L Tedalr bag), C (directly sampling of cylinder with high pressure) respectively. The result of this study showed that method C with sampling is most adequate to the proficiency testing for VOCs in indoor air.

• The Korean Journal of Applied Statistics
• /
• v.25 no.1
• /
• pp.205-213
• /
• 2012

Transforming response variable is a general tool to adapt data to a linear regression model. However, it is well known that response transformations in linear regression are very sensitive to one or a few outliers. Many methods have been suggested to develop transformations that will not be influenced by potential outliers. Recently Cheng (2005) suggested to using a trimmed likelihood estimator based on the idea of the least trimmed squares estimator(LTS). However, the method requires presetting the number of outliers and needs many computations. A new method is proposed, that can solve the problems addressed and improve the robustness of the estimates. The method uses a stepwise procedure, suggested by Hadi and Simonoff (1993), to detect outliers that determine response transformations.

• Journal of Korean Society of Steel Construction
• /
• v.10 no.1 s.34
• /
• pp.37-45
• /
• 1998

The relational database system on fatigue strength was constructed, and the properties of fatigue life distribution were examined to analyze reliability and safety of metallic materials. Data manipulations were efficiently performed in relational fatigue strength database system using dependency diagram. Regardless of the distribution of fatigue strength, the proposed method, the Robust method and the complementary error function method using probability distribution, successfully estimated parameters of the 3-parameter Weibull distribution. The proposed criterion for estimating non-failure probability showed good results regardless of censoring time. The fatigue life distribution function described as a function of parameters of the Weibull distribution and applied stress ratio produced P-S-N characteristics reasonably.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2012.05a
• /
• pp.418-418
• /
• 2012

현재 전 세계적으로 온실가스 농도 증가로 호우나 가뭄, 대설 등 지역에 따라 서로 상반되는 변화를 가져올 수 있다고 경고되고 있으며, 우리나라에서도 남해안지역과 경기북부지역에서 호우빈도가 증가하는 반면, 충정도 내륙지역과 경상북도에서는 호우빈도가 감소하고 5일 누적 강수량 또한 감소하여, 해당지역에서 가뭄이 발생할 경우 심화될 가능성이 높아진다고 보고된 바 있다. 기후변화 시나리오에 분석결과에서도 우리나라의 경우 평균적으로 강우일수는 작아지며, 강우강도는 커지는 결과들이 도출되었다. 이러한 결과들은 가뭄의 발생가능성이 높아지고 있음을 보여주고 있다. 본 연구에서는 우리나라에서 발생된 가뭄의 특성을 분석하고 가뭄의 특성과 기상인자간의 관계를 Quantile regression 분석을 통해 살펴보고자 한다. 가뭄의 특성과 기상인자(엘니뇨, 강수량 등)의 관계에 있어서 기상인자들의 평균을 이용하는 일반적인 회귀분석은 전체 데이터의 영향에 따른 가뭄특성인자와의 관계를 보여준다. 하지만 강수량과 가뭄과의 관계에서와 같이 강수량의 극값보다는 적은 강수량 혹은 무강우일수가 가뭄과 밀접한 관련을 보여준다. 이러한 점에서 이상치들에 영향을 배재할 수 있는 Quantile regression을 사용하여 Quantile에 따른 기상인자와 가뭄특성과의 관계를 규명하고 평가해 보고자 한다. 본 연구에서 적용한 Quantile Regression 기법은 회귀계수의 추정에 있어서 회귀인자의 신뢰성을 아래와 같은 Quantile-회귀계수 그래프를 통해 분석할 수 있으며, 로버스트 통계량의 특징인 분산이 적은 안정적인 추정량을 확보할 수 있는 장점을 갖는다. 아래식은 Quantile regression의 회귀계수 추정식을 나타낸다. $$arg\;in\;{n\\\;p(y_i-f(x_i,\;z_i,\;{\cdots}))\\ =1}$$ 여기서, $y_i$는 가뭄특성값을 $x_i$, $z_i$, $\cdots$는 기상인자를 나타낸다. $$p(y-q)={{\beta}(y-q)\;y{\geq_-}q \\ (1-{\beta})(q-y)\;y<q}$$ ${\beta}$는 quantile을 나타내며 0< ${\beta}$ <1범위를 갖는다.

• The Journal of Korean Society for Radiation Therapy
• /
• v.25 no.1
• /
• pp.25-31
• /
• 2013

Purpose: The aim of this study is to evaluate the dosimetric properties of robust planning strategy for plain intensity-modulated proton therapy (IMPT) taking into account of the uncertainties of effective proton range and set up error as compared to photon intensity-modulated radiation therapy (photon-IMRT) in prostate cancer treatment. Materials and Methods: The photon-IMRT (7 beams, step & shoot), plain-IMPT (2, 4, and 7 portals), and robust- IMPT plans, which was recalculated the plain-IMPT based on the uncertainties of range error (${\pm}5%$) and set up error (0.5 cm), were evaluated for five prostate cancer patients prescribed by 70 Gy/35 fractions. To quantitatively evaluate the dose distributions, several parameters such as maximum dose, minimum dose, mean dose, conformity index (CI), and homogeneity index (HI) for PTV as well as dose-volume index of VxGy for OARs were calculated from dose-volume histograms. Results: Robust-IMPT showed superior dose distributios in the PTV and OARs as compared to plain-IMPT and photon-IMRT. Like plain-IMPT, robust-IMPT were resulted in dose fluctuation around OARs, while better homogeneity and conformity in PTVs and lower mean dose in OARs as compared to photon-IMRT. Conclusion: In consideration with the effective range correction and set up movement using robustness in IMPT plan, the dosimetric uncertainties from plain-IMPT could substantially reduce and suggest more effective solutions than photon-IMRT in prostate cancer treatment.

• The Korean Journal of Applied Statistics
• /
• v.22 no.2
• /
• pp.329-340
• /
• 2009

For complex diseases such as diabetes, hypertension, it is believed that model-free methods might work better because they do not require a precise knowledge of the mode of inheritance controlling the disease trait. This is done by estimating the sharing probabilities that a pair shares zero, one, or two alleles identical by descent(IBD) and has some specific branches of test procedure, i.e., the mean test, the proportion test, and the minmax test. Among them, the minmax test is known to be more robust than others regardless of genetic mode of inheritance in current use. In this study, we compared the power of the methods which are based on minmax test and considering weighting schemes for sib-pairs to analyze sibship data. In simulation result, we found that the method based on Suarez' was more powerful than any others without respect to marker allele frequency, genetic mode of inheritance, sibship size. Also, The power of both Suarez- and Hodge-based methods was higher when marker allele frequency and sibship size were higher, and this result was remarkable in dominant mode of inheritance especially.

• The Korean Journal of Applied Statistics
• /
• v.21 no.1
• /
• pp.81-94
• /
• 2008

Family-based tests such as the transmission and disequilibrium tests(TDT) have proved to be powerful tools in the search for disease genes. Unlike case-control studies, the tests are not affected by population admixture, which can lead to spurious association of multiple highly linked makers with disease-susceptible genes. Those tests have largely required knowledge of parental marker genotypes. However, parental data are often not available for late-onset diseases. In this article we propose sib-TDTs that overcome this problem by use of marker data from unaffected sib(s) instead of parents. To do this end, we fist defined a Mantel-Haenszel-type statistic for each haplotype and then proposed two tests based on this statistic. Simulation studies suggest that the proposed tests are robust to population admixture and are monotone increasing as a relative risk increases irrespective of mode of inheritance. We also illustrated the proposed tests with data adopted from Yonsei Cardiovascular Genome Center.