• The Korean Journal of Applied Statistics
• /
• v.17 no.3
• /
• pp.489-498
• /
• 2004

In this paper, a new form of linear models referred to as generalized weighted linear models is proposed. The proposed models assume that the relationship between the response variable and explanatory variables can be modelled by a distribution function of the response mean and a weighted linear combination of distribution functions of covariates. This form addresses a structural problem of the link function in the generalized linear models in which the parameter space may not be consistent with the space derived from linear predictors. The maximum likelihood estimation with Lagrange's undetermined multipliers is used to estimate the parameters and resampling method is applied to compute confidence intervals and to test hypotheses.

• Journal of the Korean Data and Information Science Society
• /
• v.21 no.5
• /
• pp.831-839
• /
• 2010

Least squares support vector machine (LS-SVM) is a kernel trick gaining a lot of popularities in the regression and classification problems. We use LS-SVM to propose a iterative algorithm for a nonlinear generalized autoregressive conditional heteroscedasticity model in the mean (GARCH-M) model to estimate the mean and the conditional volatility of stock market returns. The proposed method combines a weighted LS-SVM for the mean and unweighted LS-SVM for the conditional volatility. In this paper, we show that nonlinear GARCH-M models have a higher performance than the linear GARCH model and the linear GARCH-M model via real data estimations.

• Journal of Korea Water Resources Association
• /
• v.42 no.12
• /
• pp.1053-1067
• /
• 2009

The objective of this study is to develop an efficient and accurate quadratic finite element model based on Streamline Upwind/Petrov Galerkin (SU/PG) scheme for analyzing and predicting two dimensional flow features in complex natural rivers. For a development of model, quadratic tin, quadrilateral and mixed elements as well as linear tin, quadrilateral and mixed elements were used in the model. Also, this model was developed through reinforcement of Gauss Quadrature which was necessary to integral of governing equation. Several tests for bottom-rising channel and U-type channel were performed for the purpose of validation and verification of the developed model. Such results showed that solutions of second order elements are better accurate and improved than those of linear elements. Results obtained by the developed model and RMA-2 model are compared, and the results for the developed model were better accurate than those of RMA-2 model. In the future if the developed model is applied in natural rivers, it can provide better accurate results than those of existing model.

• Journal of Korean Society of Transportation
• /
• v.29 no.2
• /
• pp.101-109
• /
• 2011

In most of the urbanized cities, socio-economic attributes tend to cluster as patterns of similarity in space, namely spatial autocorrelation, by agglomeration forces. The classical linear regression model, the most frequently adopted in the trip generation step, cannot sufficiently represent this effect. In order to take into account the effect properly, we need a model which adequately deals with the spatial dependence patterns. In this study, the Geographically Weighted Regression (GWR) model is adopted as an alternative method for the local analysis of relationships in multivariate data sets; that is GWR extends this traditional regression framework by estimating local rather than global parameters. This study shows the existence of spatial effects in the production and attraction of home base/non-home based trips through the GWR model using travel data collected in Daegu metropolitan area. Furthermore, LISA is employed to verify the fact that the local spatial autocorrelation exists.

• Journal of Korea Water Resources Association
• /
• v.33 no.1
• /
• pp.31-38
• /
• 2000

Relationships between hydrologic variables are often nonlinear. Usually the functional form of such a relationship is not known a priori. A multivariate, nonparametric regression methodology is provided here for approximating the underlying regression function using locally weighted polynomials. Locally weighted polynomials consider the approximation of the target function through a Taylor series expansion of the function in the neighborhood of the point of estimate. The utility of this nonparametric regression approach is demonstrated through an application to nonparametric short term forecasts of the biweekly Great Salt Lake volume.volume.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.2 no.1 s.2
• /
• pp.63-73
• /
• 2003

COSMOS(Cycle, Offset, Split Model for Seoul), a real-time traffic adaptive signal system. estimates queue lengths on each approach on the basis of arithmetic average spot speeds calculated on loop detectors installed at each of two adjacent lanes. In this paper, A new method, a traffic volume-weighted average method, was studied and compared with the existing arithmetic average method. It was found that the relationship between the ratio of volumes of two lanes and the difference of average speed of each lane has a linear form. With field data, The two methods were applied and the proposed method shows more stable and reasonable queue estimation results.

• The Korean Journal of Applied Statistics
• /
• v.34 no.6
• /
• pp.945-956
• /
• 2021

In this study, we considered the algorithm of Legendre and Legendre (2012), which derives canonical correspondence analysis from weighted principal component analysis. And, it was proved that the canonical correspondence analysis based on the weighted principal component analysis is exactly the same as Ter Braak's (1986) canonical correspondence analysis based on the Gaussian response model. Ter Braak (1986)'s canonical correspondence analysis derived from a Gaussian response curve that can explain the abundance of species in ecology well uses the basic assumption of the species packing model and then conducts generalized linear model and canonical correlation analysis. It is derived by way of binding. However, the algorithm of Legendre and Legendre (2012) is calculated in a method quite similar to Benzecri's correspondence analysis without such assumptions. Therefore, if canonical correspondence analysis based on weighted principal component analysis is used, it is possible to have some flexibility in using the results. In conclusion, this study shows that the two methods starting from different models have the same site scores, species scores, and species-environment correlations.

• Proceedings of the Korean Statistical Society Conference
• /
• 2003.10a
• /
• pp.111-115
• /
• 2003

소지역 모형들은 고정된(fixed)효과와 랜덤 효과를 포함하는 일반적 선형 혼한 모형의 특별한 경우로 간주될 수 있다. 소지역 평균이나 종계는 고정된 효과와 랜덤 효과의 일치 결합으로 표현될 수 있다. 블록 대각 공분산 구조를 갖는 선형 혼합모형(mixed model) 아래서 EBLUP은 실재문제에 있어서 많이 소지역 모형에 응용된다. 설계 가중값(design weight) 들에 의존하고 설계-일치(design consistency) 성질을 만족하는 Pseudo-EBLUP 추정량들은 소지역추정에서 합해지면 (aggregated) 사후-수정(post-adjustment)없이 벤치마킹 성질을 만족한다.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.19 no.5
• /
• pp.730-736
• /
• 2009

It is more important to analysis of huge gemonics data in Bioinformatics. Here we present a novel datamining approach to predict structure and function using protein's primnary structure only. We propose not also to develope n-Block substring search algorithm in reducing enormous search space effectively in relation to feature selection, but to formulate weighted linear algorithm in a prediction of structure and function of a protein using primary structure. And we show efficient in protein domain characterization and classification by calculation weight value in determining domain association in each selected substring, and also reveal that more efficient results are acquired through claculated model score result in an inference about degree of association with each CDS(coding sequence) in domain.

• The Korean Journal of Financial Management
• /
• v.14 no.1
• /
• pp.171-205
• /
• 1997

'80년대 중반 들어 주가지수 예측모형으로 애용되던 시계열 예측모형에 대한 근본적인 의문이 제기되었다. 이것은 기존 예측모형이 선형 데이터 생성과정을 기본가정으로 채택하고 있지만 진정한 데이터 생성과정은 비선형일 수도 있다는 점에서 출발한다. 주가지수의 변동을 유발하는 경제의 기본구조가 비선형임에도 불구하고 이를 선형모형으로 접근한다면 주가의 움직임을 제대로 설명할 수 없을 뿐만 아니라 이러한 설정오류는 모형의 신뢰성을 크게 손상시킨다. 이와 같은 점에 착안하여 본 연구는 업종별 주가지수의 비선형 검정을 통해 주가가 어떠한 형태의 경제구조에서 생성되었는지 여러 가지 방법으로 정정한다. 10개 업종지수의 검정결과 보험업을 제외한 대부분의 업종지수가 카오스 끌개를 보유하고 있다는 증거가 포착되었다. 표본외 예측을 위해서 국지적 가중회귀법을 채택하였는데 예측결과 모형에 따라 $6{\sim}7$개 업종에서 통상최소자승법보다 예측력 우위를 보였다.