• Journal of the Korean Data and Information Science Society
• /
• v.24 no.1
• /
• pp.125-133
• /
• 2013

In this paper, penalized binary logistic regression models are employed as statistical models for determining the discharge of 668 patients with a chief complaint of dyspnea based on 11 blood tests results. Specifically, the ridge model based on $L^2$ penalty and the Lasso model based on $L^1$ penalty are considered in this paper. In the comparison of prediction accuracy, our models are compared with the logistic regression models with all 11 explanatory variables and the selected variables by variable selection method. The results show that the prediction accuracy of the ridge logistic regression model is the best among 4 models based on 10-fold cross-validation.

• Journal of the Korean Data and Information Science Society
• /
• v.23 no.2
• /
• pp.385-392
• /
• 2012

Logistic regression is a well known binary classification method in the field of statistical learning. Mixed-effect regression models are widely used for the analysis of correlated data such as those found in longitudinal studies. We consider kernel extensions with semiparametric fixed effects and parametric random effects for the logistic regression. The estimation is performed through the penalized likelihood method based on kernel trick, and our focus is on the efficient computation and the effective hyperparameter selection. For the selection of optimal hyperparameters, cross-validation techniques are employed. Numerical results are then presented to indicate the performance of the proposed procedure.

• Journal of the Korean Data and Information Science Society
• /
• v.22 no.5
• /
• pp.931-940
• /
• 2011

Classification is to generate a rule of classifying objects into several categories based on the learning sample. Good classification model should classify new objects with low misclassification error. Many types of classification methods have been developed including logistic regression, discriminant analysis and tree. This paper presents a new classification method using penalized partial least squares. Penalized partial least squares can make the model more robust and remedy multicollinearity problem. This paper compares the proposed method with logistic regression and PCA based discriminant analysis by some real and artificial data. It is concluded that the new method has better power as compared with other methods.

• Communications for Statistical Applications and Methods
• /
• v.27 no.6
• /
• pp.603-624
• /
• 2020

There is an emerging interest in brain functional connectivity (FC) based on functional Magnetic Resonance Imaging in Alzheimer's disease (AD) studies. The complex and high-dimensional structure of FC makes it challenging to explore the association between altered connectivity and AD susceptibility. We develop a pipeline to refine FC as proper covariates in a penalized logistic regression model and classify normal and AD susceptible groups. Three different quantification methods are proposed for FC refinement. One of the methods is dimension reduction based on common component analysis (CCA), which is employed to address the limitations of the other methods. We applied the proposed pipeline to the Alzheimer's Disease Neuroimaging Initiative (ADNI) data and deduced pathogenic FC biomarkers associated with AD susceptibility. The refined FC biomarkers were related to brain regions for cognition, stimuli processing, and sensorimotor skills. We also demonstrated that a model using CCA performed better than others in terms of classification performance and goodness-of-fit.

• Communications for Statistical Applications and Methods
• /
• v.27 no.1
• /
• pp.129-140
• /
• 2020

In this paper, we introduce an efficient algorithm for the non-convex penalized multinomial logistic regression that can be uniformly applied to a class of non-convex penalties. The class includes most non-convex penalties such as the smoothly clipped absolute deviation, minimax concave and bridge penalties. The algorithm is developed based on the concave-convex procedure and modified local quadratic approximation algorithm. However, usual quadratic approximation may slow down computational speed since the dimension of the Hessian matrix depends on the number of categories of the output variable. For this issue, we use a uniform bound of the Hessian matrix in the quadratic approximation. The algorithm is available from the R package ncpen developed by the authors. Numerical studies via simulations and real data sets are provided for illustration.

• Communications for Statistical Applications and Methods
• /
• v.27 no.2
• /
• pp.225-239
• /
• 2020

Analysis approaches for single compositional data are well established; however, effective analysis strategies for paired compositional data remain to be investigated. The current project was motivated by studies of age-related hearing loss (presbyacusis), where subjects are classified into four audiometric phenotypes that need to be ranked within these phenotypes based on their paired compositional data. We address this challenge by formulating this problem as a classification problem and integrating a penalized multinomial logistic regression model with compositional data analysis approaches. We utilize Elastic Net for a penalty function, while considering average, absolute difference, and perturbation operators for compositional data. We applied the proposed approach to the presbyacusis study of 532 subjects with probabilities that each ear of a subject belongs to each of four presbyacusis subtypes. We further investigated the ranking of presbyacusis subjects using the proposed approach based on previous literature. The data analysis results indicate that the proposed approach is effective for ranking subjects based on paired compositional data.

• Genomics & Informatics
• /
• v.14 no.4
• /
• pp.138-148
• /
• 2016

The success of genome-wide association studies (GWASs) has enabled us to improve risk assessment and provide novel genetic variants for diagnosis, prevention, and treatment. However, most variants discovered by GWASs have been reported to have very small effect sizes on complex human diseases, which has been a big hurdle in building risk prediction models. Recently, many statistical approaches based on penalized regression have been developed to solve the "large p and small n" problem. In this report, we evaluated the performance of several statistical methods for predicting a binary trait: stepwise logistic regression (SLR), least absolute shrinkage and selection operator (LASSO), and Elastic-Net (EN). We first built a prediction model by combining variable selection and prediction methods for type 2 diabetes using Affymetrix Genome-Wide Human SNP Array 5.0 from the Korean Association Resource project. We assessed the risk prediction performance using area under the receiver operating characteristic curve (AUC) for the internal and external validation datasets. In the internal validation, SLR-LASSO and SLR-EN tended to yield more accurate predictions than other combinations. During the external validation, the SLR-SLR and SLR-EN combinations achieved the highest AUC of 0.726. We propose these combinations as a potentially powerful risk prediction model for type 2 diabetes.

• Genomics & Informatics
• /
• v.17 no.4
• /
• pp.47.1-47.12
• /
• 2019

The achievements of genome-wide association studies have suggested ways to predict diseases, such as type 2 diabetes (T2D), using single-nucleotide polymorphisms (SNPs). Most T2D risk prediction models have used SNPs in combination with demographic variables. However, it is difficult to evaluate the pure additive contribution of genetic variants to classically used demographic models. Since prediction models include some heritable traits, such as body mass index, the contribution of SNPs using unmatched case-control samples may be underestimated. In this article, we propose a method that uses propensity score matching to avoid underestimation by matching case and control samples, thereby determining the pure additive contribution of SNPs. To illustrate the proposed propensity score matching method, we used SNP data from the Korea Association Resources project and reported SNPs from the genome-wide association study catalog. We selected various SNP sets via stepwise logistic regression (SLR), least absolute shrinkage and selection operator (LASSO), and the elastic-net (EN) algorithm. Using these SNP sets, we made predictions using SLR, LASSO, and EN as logistic regression modeling techniques. The accuracy of the predictions was compared in terms of area under the receiver operating characteristic curve (AUC). The contribution of SNPs to T2D was evaluated by the difference in the AUC between models using only demographic variables and models that included the SNPs. The largest difference among our models showed that the AUC of the model using genetic variants with demographic variables could be 0.107 higher than that of the corresponding model using only demographic variables.

• Journal of Korean Neurosurgical Society
• /
• v.66 no.5
• /
• pp.536-542
• /
• 2023

Objective : Surgical site infection is the most detrimental complication following cranioplasty. In other surgical fields, intrawound vancomycin powder application has been introduced to prevent surgical site infection and is widely used based on results in multiple studies. This study evaluated the effect of intrawound vancomycin powder in cranioplasty compared with the conventional method without topical antibiotics. Methods : This retrospective study included 580 patients with skull defects who underwent cranioplasty between August 1, 1998 and December 31, 2021. The conventional method was used in 475 (81.9%; conventional group) and vancomycin powder (1 g) was applied on the dura mater and bone flap in 105 patients (18.1%; vancomycin powder group). Surgical site infection was defined as infection of the incision, organ, or space that occurred after cranioplasty. Surgical site infection within 1-year surveillance period was compared between the conventional and vancomycin powder groups with logistic regression analysis. Penalized likelihood estimation method was used in logistic regression to deal with zero events. All local and systemic adverse events associated with topical vancomycin application were also evaluated. Results : Surgical site infection occurred in 31 patients (5.3%) and all were observed in the conventional group. The median time between cranioplasty and detection of surgical site infection was 13 days (range, 4-333). Staphylococci were the most common organisms and identified in 25 (80.6%) of 31 cases with surgical site infections. The surgical site infection rate in the vancomycin powder group (0/105, 0.0%) was significantly lower than that in the conventional group (31/475, 6.5%; crude odds ratio [OR], 0.067; 95% confidence interval [CI], 0.006-0.762; adjusted OR, 0.068; 95% CI, 0.006-0.731; p=0.026). No adverse events associated with intrawound vancomycin powder were observed during the follow-up. Conclusion : Intrawound vancomycin powder effectively prevented surgical site infections following cranioplasty without local or systemic adverse events. Our results suggest that intrawound vancomycin powder is an effective and safe strategy for patients undergoing cranioplasty.

• Neonatal Medicine
• /
• v.25 no.4
• /
• pp.161-169
• /
• 2018

Purpose: We assessed the influence of antenatal corticosteroid (ACS) on the inhospital outcomes of intrauterine growth restriction (IUGR) infants. Methods: A retrospective study was conducted with singletons born at $23^{+0}$ to $33^{+6}weeks$ of gestation at Seoul National University Hospital from 2007 to 2014. We compared clinical outcomes between infants who received ACS 2 to 7 days before birth (complete ACS), at <2 or >7 days (incomplete ACS), and those who did not receive ACS in IUGR and AGA infants. Multivariate logistic regression using Firth's penalized likelihood was performed. Results: 304 neonates with 91 IUGR neonates were eligible. Among AGA neonates, mortality (adjusted odds ratio [aOR], 0.13; 95% confidence interval [CI], 0.02 to 0.78), hypotension within 7 postnatal days (aOR, 0.20; 95% CI, 0.06 to 0.64), and severe bronchopulmonary dysplasia (BPD) or death (aOR, 0.24; 95% CI, 0.07 to 0.77) were lower in complete ACS group after adjusting for pregnancy induced hypertension and uncontrolled preterm labor. Mortality (aOR, 0.18; 95% CI, 0.04 to 0.78), hypotension (aOR, 0.26; 95% CI, 0.09 to 0.70), and severe BPD or death (aOR, 0.33; 95% CI, 0.12 to 0.92) were also lower in the incomplete ACS group. Among IUGR infants, after adjusting for birth weight and 5-minute Apgar score, inhaled nitric oxide use within 14 postnatal days was lower in both complete ACS (aOR, 0.07; 95% CI, 0.01 to 0.67) and incomplete ACS (aOR, 0.04; 95% CI, 0.01 to 0.37) groups. Conclusion: ACS was not effective in reducing morbidities in IUGR preterm infants.