• Journal of the Korean Data and Information Science Society
• /
• v.20 no.6
• /
• pp.1093-1101
• /
• 2009

Quantile regression has become a more widely used technique to describe the distribution of a response variable given a set of explanatory variables. This paper proposes a novel modelfor quantile regression using doubly penalized kernel machine with support vector machine iteratively reweighted least squares (SVM-IRWLS). To make inference about the shape of a population distribution, the widely popularregression, would be inadequate, if the distribution is not approximately Gaussian. We present a likelihood-based approach to the estimation of the regression quantiles that uses the asymmetric Laplace density.

• The Korean Journal of Applied Statistics
• /
• v.28 no.5
• /
• pp.965-976
• /
• 2015

Determining relevant variables for a regression model is important in regression analysis. Recently, a variable selection methods using a penalized likelihood with various penalty functions (e.g. LASSO and SCAD) have been widely studied in simple statistical models such as linear models and generalized linear models. The advantage of these methods is that they select important variables and estimate regression coefficients, simultaneously; therefore, they delete insignificant variables by estimating their coefficients as zero. We study how to select proper variables based on penalized hierarchical likelihood (HL) in semi-parametric frailty models that allow three penalty functions, LASSO, SCAD and HL. For the variable selection we develop a new function in the "frailtyHL" R package. Our methods are illustrated with breast cancer survival data from the Medical Center at Chonnam National University in Korea. We compare the results from three variable-selection methods and discuss advantages and disadvantages.

• The Korean Journal of Applied Statistics
• /
• v.28 no.6
• /
• pp.1121-1131
• /
• 2015

Logistic models with a random intercept are useful to analyze longitudinal binary data. Traditionally, the random intercept of the logistic model is assumed to be parametric (such as normal distribution) and is also assumed to be independent to variables. Such assumptions are very strong and restricted for application to real data. Recently, Garcia and Ma (2015) derived semiparametric efficient estimators for logistic model with a random intercept without these assumptions. Their estimator shows the consistency where we do not assume any parametric form for the random intercept. In addition, the method is computationally simple. In this paper, we apply this method to analyze toenail infection data. We compare the semiparametric estimator with maximum likelihood estimator, penalized quasi-likelihood estimator and hierarchical generalized linear estimator.

• Journal of the Korean Society of Safety
• /
• v.32 no.3
• /
• pp.141-150
• /
• 2017

A rapid increase in traffic accidents involving senior vehicle occupants has been an issue in Korea because of the aging of the population occurring at one of the fastest rates in the world; unfortunately, few studies beyond several looking at the effect of senior occupants on the level of accident injury severity can be found in the literature. A Multinomial logit model was estimated with Newton-Raphson algorithm to perform bias-reducing penalized likelihood optimization. Model covariates integral to developing the model were included, but the main focus was on the interaction of seating position and injury to senior vehicle occupants. It was found that the likelihood of an accident resulting in a fatality increased: 2.2 times for the driver seat, 2.7 times for the front passenger seat, and even 6.7 times for the rear seat. A mandatory seatbelt law to be extended to the rear seat needs to pass the assembly as soon as possible, and government, industry, and safety groups should be encouraged to join forces to strongly carry out targeted campaigns for the wearing of seatbelts in all vehicle seats to enhance the safety of senior occupants as well as other occupants who are vulnerable to road traffic accidents.

• Journal of the Korean Data and Information Science Society
• /
• v.19 no.4
• /
• pp.1419-1427
• /
• 2008

This paper shows the kernel Poisson regression which can be applied in the claims reserving, where the row effect is assumed to be a nonlinear function of the row index. The paper concentrates on the chain-ladder technique, within the framework of the chain-ladder linear model. It is shown that the proposed method can provide better reserve estimates than the Poisson model. The cross validation function is introduced to choose optimal hyper-parameters in the procedure. Experimental results are then presented which indicate the performance of the proposed model.

• Journal of the Korean Data and Information Science Society
• /
• v.19 no.4
• /
• pp.1353-1360
• /
• 2008

An estimating procedure is introduced for the nonlinear mixed-effect Poisson regression, for longitudinal study, where data from different subjects are independent whereas data from same subject are correlated. The proposed procedure provides the estimates of the mean function of the response variables, where the canonical parameter is related to the input vector in a nonlinear form. The generalized cross validation function is introduced to choose optimal hyper-parameters in the procedure. Experimental results are then presented, which indicate the performance of the proposed estimating procedure.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.12
• /
• pp.3694-3702
• /
• 1999

OSL algorithm has an advantage that repeated algorithm is easily derived, even though penalty function which has a complicated transcendental function. In order to solve this problem, we suggested MPEMG algorithm. However, though this algorithm extend convergence rate of smoothing constant, it include the problem that is not faster than OSL algorithm in the convergence rate increasing penalized log-likelihood. Accordingly, in this paper, we will suggest SMOSLG digital image restoration algorithm which is fast in the convergence rate as well as extend convergence region of smoothing constant. And also we will study the usefulness of algorithm suggested through digital image simulation.

• Communications for Statistical Applications and Methods
• /
• v.26 no.1
• /
• pp.69-78
• /
• 2019

A structured model with both single-index and varying coefficients is a powerful tool in modeling high dimensional data. It has been widely used because the single-index can overcome the curse of dimensionality and varying coefficients can allow nonlinear interaction effects in the model. For high dimensional index vectors, variable selection becomes an important question in the model building process. In this paper, we propose an efficient estimation and a variable selection method based on a smoothing spline approach in a partially linear single-index-coefficient regression model. We also propose an efficient algorithm for simultaneously estimating the coefficient functions in a data-adaptive lower-dimensional approximation space and selecting significant variables in the index with the adaptive LASSO penalty. The empirical performance of the proposed method is illustrated with simulated and real data examples.

• Genomics & Informatics
• /
• v.14 no.4
• /
• pp.187-195
• /
• 2016

In genetic association studies with high-dimensional genomic data, multiple group testing procedures are often required in order to identify disease/trait-related genes or genetic regions, where multiple genetic sites or variants are located within the same gene or genetic region. However, statistical testing procedures based on an individual test suffer from multiple testing issues such as the control of family-wise error rate and dependent tests. Moreover, detecting only a few of genes associated with a phenotype outcome among tens of thousands of genes is of main interest in genetic association studies. In this reason regularization procedures, where a phenotype outcome regresses on all genomic markers and then regression coefficients are estimated based on a penalized likelihood, have been considered as a good alternative approach to analysis of high-dimensional genomic data. But, selection performance of regularization procedures has been rarely compared with that of statistical group testing procedures. In this article, we performed extensive simulation studies where commonly used group testing procedures such as principal component analysis, Hotelling's $T^2$ test, and permutation test are compared with group lasso (least absolute selection and shrinkage operator) in terms of true positive selection. Also, we applied all methods considered in simulation studies to identify genes associated with ovarian cancer from over 20,000 genetic sites generated from Illumina Infinium HumanMethylation27K Beadchip. We found a big discrepancy of selected genes between multiple group testing procedures and group lasso.

• The Korean Journal of Applied Statistics
• /
• v.24 no.6
• /
• pp.1213-1224
• /
• 2011

In high dimensionality where the number of variables are excessively larger than observations, it is required to remove the noninformative variables to cluster observations. Most model-based approaches for variable selection have been considered under the assumption of homoscedasticity and their models are mainly estimated by a penalized likelihood method. In this paper, a different approach is proposed to remove the noninformative variables effectively and to cluster based on the modified normal mixture model simultaneously. The validity of the model was provided and an EM algorithm was derived to estimate the parameters. Simulation studies and an experiment using real microarray dataset showed the effectiveness of the proposed method.