• Communications for Statistical Applications and Methods
• /
• v.23 no.4
• /
• pp.355-362
• /
• 2016

Ensemble methods often help increase prediction ability in various predictive models by combining multiple weak learners and reducing the variability of the final predictive model. In this work, we demonstrate that ensemble methods also enhance the accuracy of prediction under kernel ridge regression and kernel logistic regression classification. Here we apply bagging and random forests to two kernel-based predictive models; and present the procedure of how bagging and random forests can be embedded in kernel-based predictive models. Our proposals are tested under numerous synthetic and real datasets; subsequently, they are compared with plain kernel-based predictive models and their subsampling approach. Numerical studies demonstrate that ensemble approach outperforms plain kernel-based predictive models.

• Asian Pacific Journal of Cancer Prevention
• /
• v.16 no.2
• /
• pp.647-652
• /
• 2015

Background: Lung cancer was one of the most common cancers in both men and women all over the world. In this study, we aimed to clarify who could survive after long term chemotherapy in patients with advanced non-small cell lung cancer (NSCLC). Methods: We enrolled 186 patients with stage IV NSCLC after long term chemotherapy from Jun 2006 to Nov 2014 diagnosed in Jiangsu Cancer Hospital. Multiple variables like age, gender, smoking, histology of adenocarcinoma and squamous-cell cancer, number of metastatic sites, metastatic sites (e.g. lung, brain, bone, liver and pleura), hemoglobin, lymphocyte rate (LYR), Change of LYR during multiple therapies, hypertension, diabetes, chronic bronchitis, treatments (e.g.radiotherapy and targeted therapy) were selected. For consideration of factors influencing survival and response for patients with advanced NSCLC, logistic regression analysis and Cox regression analysis were used in an attempt to develop a screening module for patients with elevated survival after long term chemotherapy become possible. Results: Of the total of 186 patients enrolled, 69 survived less than 1 year (short-term group), 45 one to two years, and 72 longer than 3 years (long-term group). For logistic regression analysis, the short-term group was taken as control group and the long-term group as the case group. We found that age, histology of adenocarcinoma, metastatic site (e.g. lung and liver), treatments (e.g. targeted therapy and radiotherapy), LYR, a decreasing tendency of LYR and chronic bronchitis were individually associated with overall survival by Cox regression analysis. A multivariable Cox regression model showed that metastatic site (e.g. lung and liver), histology of adenocarcinoma, treatments (e.g. targeted therapy and radiotherapy) and chronic bronchitis were associated with overall survival. Thus metastatic site (e.g. lung and liver) and chronic bronchitis may be important risk factors for patients with advanced NSCLC. Gender, metastatic site (e.g. lung and liver), LYR and the decreasing tendency of LYR were significantly associated with long-term survival in the individual-variable logistic regression model (P<0.05). On multivariate logistic regression analysis, gender, metastatic site (e.g. lung and liver) and the decreasing tendency of LYR associated with long-term survival. Conclusions: In conclusion, female patients with stage IV adenocarcinoma of NSCLC who had decreasing tendency of LYR during the course therapy and had accepted multiple therapies e.g. more than third-line chemotherapy, radiotherapy and/or targeted therapy might be expected to live longer.

• The Journal of Korean Academy of Orthopedic Manual Physical Therapy
• /
• v.18 no.1
• /
• pp.19-29
• /
• 2012

Background: This study was conducted to investigate subjective musculoskeletal symptom and the related factor of caregiver. Methods: For 300 caregiver, we used the self-administered questionnaires to examine occupational stress and subjective musculoskeletal symptom designed by KOSHA. The collected data were analyzed chi-square test, independent t-test and multiple logistic regression analysis using SPSS 12.0. Results: The multiple logistic regression analysis showed that the caregiver working in the general hospital significantly increased the subjective musculoskeletal symptom in their neck, shoulder, hand/wrist/finger, back, leg/foot. For the caregiver working in hospital showed significantly increased the subjective musculoskeletal symptom in their hand/wrist/finger and leg/foot. Conclusions: With the above results, continuous and systematic prevention program should be established, which include the ergonomics and psychosocial factor for the caregiver's musculoskeletal symptom.

• Korean Journal of Health Education and Promotion
• /
• v.30 no.3
• /
• pp.25-34
• /
• 2013

Objectives: The main purpose of this study was to examine factors influencing school bullying victimization of middle school students in relation to social support, self-esteem, stress coping behavior, and health status. Methods: The questionnaire survey was carried out on a convenience sample of 441 middle school students. The data analysis procedure included frequency, ${\chi}^2$-test, t-test, and multiple logistic regression. Results: It was found that 18% of the subjects were bullied by other students. Multiple logistic regression analysis revealed that the factors such as sex(OR=2.35, p=.006), aggressive coping behavior(OR=1.18, p=.028), and health status(OR=1.04, p=.002) were significant affecting factors. Conclusions: The findings suggest that to prevent middle school students' bullying victimization, it is necessary to design intervention programs that considering their health status and stress coping behavior.

• Women's Health Nursing
• /
• v.22 no.3
• /
• pp.162-169
• /
• 2016

Purpose: This study was done to investigate the prevalence and factors affecting body shape misperception among Korean adult women. Methods: We analyzed raw data from the fourth Korean Longitudinal Survey of Women and Families, for 7,658 women over 20 years of age. For the multiple logistic regression model, the dependent variable was existence of misperception (yes/no). Results: The prevalence of body shape misperception was 42.6% (overestimation 6.5%, underestimation 36.1%). Multiple logistic regression analysis revealed that age, education level, marital status, economic status and BMI were significantly associated with body shape underestimation. Education level and BMI were significantly associated with body shape overestimation. Conclusion: An effective obesity intervention program for women must reflect and monitor body shape perception as well as BMI.

• Journal of East-West Nursing Research
• /
• v.16 no.2
• /
• pp.156-163
• /
• 2010

Purpose: The purpose of this study was to investigate factors associated with childhood obesity. Methods: In this descriptive study, the number of participants was 78 4-6th grade students in a primary school, and the subjects were overweight or obese. Data were collected for 10 days from June 1 to 10 in 2009. The data were analyzed using descriptive statistics, independent-samples t-test, $X^2$-test, and multiple logistic regression. Results: Mother's occupation is a statistically significant factor in a bivariate analysis. In multiple logistic regression analysis, the statistically significant factors were mother's occupation (OR=0.17, 95% CI=0.04-0.70) and dietary attitudes (OR=0.08, 95% CI=0.01-0.66). Conclusion: For an effective obesity management program in childhood, it is necessary to consider a mother occupation and dietary attitudes.

• Journal of Korean Academy of Nursing
• /
• v.47 no.3
• /
• pp.357-366
• /
• 2017

Purpose: The aim of this study was to determine the influence of various factors on preterm labor symptoms (PLS) and preterm birth (PB). Methods: This prospective cohort study included 193 women in the second stage of pregnancy. Multiple characteristics including body mass index (BMI), smoking, and pregnancy complications were collected through a self-report questionnaire. Pregnancy stress and PLS were each measured with a related scale. Cervical length and birth outcome were evaluated from medical charts. Multiple regression was used to predict PLS and logistic regression was used to predict PB. Results: Multiple regression showed smoking experience, pregnancy complications and pregnancy specific stress were predictors of PLS and accounted for 19.2% of the total variation. Logistic regression showed predictors of PB to be twins (OR=13.68, CI=3.72~50.33, p<.001), shorter cervix (<25mm) (OR=5.63, CI=1.29~24.54, p<.05), BMI >25 ($kg/m^2$) (OR=3.50, CI=1.35~9.04, p<.01) and a previous PB (OR=4.15, CI=1.07~16.03, p<.05). Conclusion: The results of this study show that the multiple factors affect stage II pregnant women can result in PLS or PB. And preterm labor may predict PB. These findings highlight differences in predicting variables for pretrm labor and for PB. Future research is needed to develop a screening tool to predict the risk of preterm birth in pregnant women.

• Health Policy and Management
• /
• v.30 no.3
• /
• pp.399-408
• /
• 2020

Background: The purpose of this study was to examine the impact of the regional characteristics on the accessibility of emergency care and the impact of emergency medical accessibility on the patients' prognosis and the emergency medical expenditure. Methods: This study used the 13th beta version 1.6 annual data of Korea Health Panel and the statistics from the Korean Statistical Information Service. The sample included 8,119 patients who visited the emergency centers between year 2013 and 2017. The arrival time, which indicated medical access, was used as dependent variable for multi-level analysis. For ordinal logistic regression and multiple regression, the arrival time was used as independent variable while patients' prognosis and emergency medical expenditure were used as dependent variables. Results: The results for the multi-level analysis in both the individual and regional variables showed that as the number of emergency medical institutions per 100 km² area increased, the time required to reach emergency centers significantly decreased. Ordinal logistic regression and multiple regression results showed that as the arrival time increased, the patients' prognosis significantly worsened and the emergency medical expenses significantly increased. Conclusion: In conclusion, the access to emergency care was affected by regional characteristics and affected patient outcomes and emergency medical expenditure.

• Journal of Institute of Control, Robotics and Systems
• /
• v.20 no.4
• /
• pp.428-436
• /
• 2014

This paper proposes a statistical regression method for classifying pillars and vehicles in parking area using a single ultrasonic sensor. There are three types of information provided by the ultrasonic sensor: TOF, the peak and the width of a pulse, from which 67 different features are extracted through segmentation and data preprocessing. The classification using the multiple SVM and the multinomial logistic regression are applied to the set of extracted features, and has achieved the accuracy of 85% and 89.67%, respectively, over a set of real-world data. The experimental result proves that the proposed feature extraction and classification scheme is applicable to the object classification using an ultrasonic sensor.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.62 no.5
• /
• pp.1-13
• /
• 2020

The empty houses' problem is important in the local revitalization and local sustainability, and these phenomenon caused by various factors of the region. The population and housing census data are the most effective data available to study this phenomenon by small regions. In this study, logistic regression and multiple regression analysis were performed to understand the effects of population, household, and housing characteristics on empty houses using population and housing census data. Also, the scale and direction of the effect of each characteristic in large cities, small cities, and rural areas were compared. As results, there was a slight difference between cities and province regions in the district and housing characteristic variables. In the comparison of Eup-Myeon-Dong, the affected variables were different in the Dong and Myeon areas. The significance of this study is to examine the effect of the characteristics of population and housing on the vacant houses and to confirm that the factors affecting different regions.