• The Korean Journal of Applied Statistics
• /
• v.34 no.5
• /
• pp.697-710
• /
• 2021

Relative error prediction is preferred over ordinary prediction methods when relative/percentile errors are regarded as important, especially in econometrics, software engineering and government official statistics. The relative error prediction techniques have been developed in linear/nonlinear regression, nonparametric regression using kernel regression smoother, and stationary time series models. However, random effect models have not been used in relative error prediction. The purpose of this article is to extend relative error prediction to some of generalized linear mixed model (GLMM) with panel data, which is the random effect models based on gamma, lognormal, or inverse gaussian distribution. For better understanding, the real auto insurance data is used to predict the claim size, and the best predictor and the best relative error predictor are comparatively illustrated.

• Communications for Statistical Applications and Methods
• /
• v.15 no.6
• /
• pp.993-1002
• /
• 2008

Correlated responses have been widely analyzed since Liang and Zeger (1986) introduced the famous Generalized Estimating Equations(GEE). However, their variance functions were restricted to known quantifies multiplied by scale parameter. In so many industries and academic/research fields, power-of-the-mean variance function is one of the common variance function. We suggest GEE-type pseudolikelihood estimation based on the power-of-the-mean variance using existing software and investigate it's efficiency for different working correlation matrices.

• 제어로봇시스템학회:학술대회논문집
• /
• 1992.10b
• /
• pp.575-580
• /
• 1992

Mixed H$_{2}$/H$_{\infty}$ robust control synthesis is considered for finite dimensional linear time-invariant systems under the presence of diagonal structured uncertainties. Such uncertainties arise for instance when there is real perturbation in the nominal model of the state space system or when modeling multiple (unstructured) uncertainty at different locations in the feedback loop. This synthesis problem is reduced to convex optimization problem over a bounded subset of matrices as well as diagonal matrix having certain structure. For computational purpose, this convex optimization problem is further reduced into Generalized Eigenvalue Minimization Problem where a powerful algorithm based on interior point method has been recently developed..

• Radiation Oncology Journal
• /
• v.34 no.2
• /
• pp.135-144
• /
• 2016

Purpose: The purpose of this study was to evaluate the efficacy of art therapy to control fatigue in cancer patients during course of radiotherapy and its impact on quality of life (QoL). Materials and Methods: Fifty cancer patients receiving radiotherapy received weekly art therapy sessions using famous painting appreciation. Fatigue and QoL were assessed using the Brief Fatigue Inventory (BFI) Scale and the Functional Assessment of Chronic Illness Therapy-Fatigue (FACIT-F) at baseline before starting radiotherapy, every week for 4 weeks during radiotherapy, and at the end of radiotherapy. Mean changes of scores over time were analyzed using a generalized linear mixed model. Results: Of the 50 patients, 34 (68%) participated in 4 sessions of art therapy. Generalized linear mixed models testing for the effect of time on mean score changes showed no significant changes in scores from baseline for the BFI and FACIT-F. The mean BFI score and FACIT-F total score changed from 3.1 to 2.7 and from 110.7 to 109.2, respectively. Art therapy based on the appreciation of famous paintings led to increases in self-esteem by increasing self-realization and forming social relationships. Conclusion: Fatigue and QoL in cancer patients with art therapy do not deteriorate during a period of radiotherapy. Despite the single-arm small number of participants and pilot design, this study provides a strong initial demonstration that art therapy of appreciation for famous painting is worthy of further study for fatigue and QoL improvement. Further, it can play an important role in routine practice in cancer patients during radiotherapy.

• Communications for Statistical Applications and Methods
• /
• v.26 no.3
• /
• pp.315-323
• /
• 2019

Panel data sets have recently been developed in various areas, and many recent studies have analyzed panel, or longitudinal data sets. Often a dichotomous dependent variable occur in survival analysis, biomedical and epidemiological studies that is analyzed by a generalized linear mixed effects model (GLMM). The most common estimation method for the binary panel data may be the maximum likelihood (ML). Many statistical packages provide ML estimates; however, the estimates are computed from numerically approximated likelihood function. For instance, R packages, pglm (Croissant, 2017) approximate the likelihood function by the Gauss-Hermite quadratures, while Rchoice (Sarrias, Journal of Statistical Software, 74, 1-31, 2016) use a Monte Carlo integration method for the approximation. As a result, it can be observed that different packages give different results because of different numerical computation methods. In this note, we discuss the pros and cons of numerical methods compared with the exact computation method.

• Journal of Preventive Medicine and Public Health
• /
• v.50 no.6
• /
• pp.377-385
• /
• 2017

Objectives: Although mercury (Hg) exposure is known to be neurotoxic in humans, its effects on liver function have been less often reported. The aim of this study was to investigate whether total Hg exposure in Korean adults was associated with elevated serum levels of the liver enzymes aspartate aminotransferase (AST), alanine transaminase (ALT), and gamma-glutamyltransferase (GGT). Methods: We repeatedly examined the levels of total Hg and liver enzymes in the blood of 508 adults during 2010-2011 and 2014-2015. Cross-sectional associations between levels of blood Hg and liver enzymes were analyzed using a generalized linear model, and nonlinear relationships were analyzed using a generalized additive mixed model. Generalized estimating equations were applied to examine longitudinal associations, considering the correlations of individuals measured repeatedly. Results: GGT increased by 11.0% (95% confidence interval [CI], 4.5 to 18.0%) in women and 8.1% (95% CI, -0.5 to 17.4%) in men per doubling of Hg levels, but AST and ALT were not significantly associated with Hg in either men or women. In women who drank more than 2 or 3 times per week, AST, ALT, and GGT levels increased by 10.6% (95% CI, 4.2 to 17.5%), 7.7% (95% CI, 1.1 to 14.7%), and 37.5% (95% CI,15.2 to 64.3%) per doubling of Hg levels, respectively, showing an interaction between blood Hg levels and drinking. Conclusions: Hg exposure was associated with an elevated serum concentration of GGT. Especially in women who were frequent drinkers, AST, ALT, and GGT showed a significant increase, with a significant synergistic effect of Hg and alcohol consumption.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.56 no.10
• /
• pp.1721-1730
• /
• 2007

SThis paper proposes a new dispatch scheduling algorithm in interconnected regional system operations. The dispatch scheduling formulated as mixed integer non-linear programming (MINLP) problem can efficiently be computed by generalized Benders decomposition (GBD) algorithm. GBD guarantees adequate computation speed and solution convergency since it decomposes a primal problem into a master problem and subproblems for simplicity. In addition, the inter-temporal optimal power flow (OPF) subproblem of the dispatch scheduling problem is comprised of various variables and constraints considering time-continuity and it makes the inter-temporal OPF complex due to increased dimensions of the optimization problem. In this paper, regional decomposition technique based on auxiliary problem principle (APP) algorithm is introduced to obtain efficient inter-temporal OPF solution through the parallel implementation. In addition, it can find the most economic dispatch schedule incorporating power transaction without private information open. Therefore, it can be expanded as an efficient dispatch scheduling model for interconnected system operation.

• The Korean Journal of Applied Statistics
• /
• v.28 no.2
• /
• pp.269-279
• /
• 2015

This paper considers inferences of random effects. We show that the proposed confidence distribution (CD) performs well in logistic regression for random intercepts with small samples. Real data analyses are also done to identify the subject effects clearly.

• Journal of the Korean Data and Information Science Society
• /
• v.27 no.5
• /
• pp.1169-1182
• /
• 2016

Cancer is a typical cause of death in Korea that becomes a major issue in health care. According to Cause of Death Statistics (2014) by National Statistical Office, SMRs (standardized mortality rates) in Busan were counted as the highest among all cities. In this paper, we used data of Busan Regional Cancer Center to estimate the extent of the cancer incidence rate and cancer mortality rate. The data are considered in small areas of administrative units such as Gu/Dong from years 2003 to 2009. All cancer including four major cancers (stomach cancer, colorectal cancer, lung cancer, liver cancer) have been analyzed. We carried out model selection and parameter estimation using spatial multi-level model incorporating a spatial correlation. For the spatial effects, CAR (conditional autoregressive model) has been assumed.

• Korean Journal of Human Ecology
• /
• v.22 no.1
• /
• pp.117-129
• /
• 2013

The purpose of this study is to examine the factors affecting on CRPI (capability of retirement planning and implementing) and to identify the critical point of financial factors to increase CRPI. 908 Korean individuals' data were used for descriptive statistics, t-test, F-test, and Generalized Linear Mixed Model Analysis. As a result, CRPI was influenced by attitude toward retirement planning, gender, income, and propensity to save. Some mid-group did not make a significant difference in CRPI compared to the lower-group. Men were more capable at CRPI than women in terms of sense of independence and getting a consulting service. Most consumers appear to have difficulty in calculating their retirement fund. People who have income over 3,000,000 won have dramatically greater CRPI than people who have income under 3,000,000 won. And people who have saving ratio over 23% have significantly greater CRPI than people who have propensity to save under 23%. The monthly income 3,000,000 won and propensity to save 23% were considered as critical points related to CRPI.