• Health Policy and Management
• /
• v.23 no.4
• /
• pp.358-368
• /
• 2013

Background: There are some data to support the theory of a protective effect of parenthood against suicide, as proposed by Durkheim in 1897. We conducted this study to examine the relationship between children characteristics and suicidal ideation among Korean married women. Methods: The data were obtained from the Korean National Health and Nutrition Examination Surveys IV, conducted in 2007-2009, respectively (3,202 married women aged 20-49 years). The children characteristics are the age of the youngest child and the number of children living with subject. Data were analyzed by using logistic regression with SAS 9.2. Results: The risk for suicidal ideation was higher among married women having a young child (aged <7 years). It had strong effect on suicidal ideation in non-employee subjects. The odds ratio for suicidal ideation of women having a youngest child aged 2-3 years compared with reference groups was 1.673 with statistical significance. But the association of suicidal ideation with number of children shows insignificant. Conslusion: The findings suggest a different result with the theory, the protective effect of being a parent of young child on suicide risk for women, first suggested by Durkheim and supported by previous studies and the need for further prospective investigation that lead to policies according children characteristics aimed at improving married women's life.

• The Korean Journal of Applied Statistics
• /
• v.30 no.5
• /
• pp.681-690
• /
• 2017

Various imbalanced binary classification problems exist such as fraud detection in banking operations, detecting spam mail and predicting defective products. Several sampling methods such as over sampling, under sampling, SMOTE have been developed to overcome the poor prediction performance of binary classifiers when the proportion of one group is dominant. In order to overcome this problem, several sampling methods such as over-sampling, under-sampling, SMOTE have been developed. In this study, we investigate prediction performance of logistic regression, Lasso, random forest, boosting and support vector machine in combination with the sampling methods for binary imbalanced data. Four real data sets are analyzed to see if there is a substantial improvement in prediction performance. We also emphasize some precautions when the sampling methods are implemented.

• Journal of Technology Innovation
• /
• v.18 no.1
• /
• pp.21-51
• /
• 2010

This paper examines the knowledge transfer effect in Korean venture systems. Existing literature has provided rich evidence of the effect of knowledge transfer, but we do not have micro mechanisms inherent in the process of knowledge transfer. This paper argues that knowledge transfer effects vary depending on the knowledge types, sources, and legacy. This paper also tests role of the two important pillars in knowledge transfer of Korean venture startups; venture capital and government. This paper also examines the role of absorptive capacity in the knowledge transfer process. With 1,862 sample of Korean venture firms, this study employed three methods depending on 3 different types of dependent variables: hierarchical regression, logistic regression, and survival analysis. Main findings include that 1) knowledge characteristic itself and its alignment with industry influence the knowledge transfer effects, 2) government support has a negative effect on financial performance of venture firms, but does not have significant interaction effect on knowledge transfer, and 3) the absorptive capacity of each firm moderates the knowledge transfer effects. The theoretical and practical implications are discussed.

• The Journal of Korean Society for School & Community Health Education
• /
• v.12 no.1
• /
• pp.1-20
• /
• 2011

Purpose: The purpose of this study is to analyze factors affecting suicidal ideation of athletes in middle and high school. Methods: The data in this study came from the survey which conducted to 500 students of physical education high school athletes and elite athletes in general school from the beginning of December 2009 to the end of January 2010 in Seoul and Kyunggi. Finally 436 cases was used. The dependant variable is suicidal ideation in this study. And the independent variables are socio-demographic factors, exercise-related factors, school and family environments, stress and depression factor. The analysis method is logistic regression analysis. Finally 436 cases was used. Results: First, the rate of suicidal ideation of athletes in middle and high school was 28.9%. Second, there were significant differences between stress and sex, grades, athletic career, kind of sport, sport skill level, the average exercise time per day, family circumstances, life satisfaction, future expectations. Third, there were significant differences between depression and sex, academic achievements, the average exercise time per day, family circumstances, life satisfaction, future expectations. Fourth, there were significant differences between suicidal ideation and sport skill level, academic achievements, the average exercise time per day, family circumstances, life satisfaction, future expectations. Fifth, the higher level of stress and degree of depression, the more increased the risk of suicidal ideation. Conclusions: Considering the close relationship between suicidal ideation and suicidal attempt, to reduce the suicidal ideation of athletes in middle and high school it is necessary to adjust exercising time properly and to promote life satisfaction. Also to overcome depression and stress support system should be established and to promote self-efficacy cognition strength program should be introduced in the school. So athletes in middle and high school should have confidence and self-confidence on the future.

• Health Policy and Management
• /
• v.10 no.3
• /
• pp.1-18
• /
• 2000

This study was conducted to find medical care utilization pattern and to examine the affecting factors on medical facilities utilization using Andersen's medical care service behavioral model. Three hundreds and five public officials with detected disease through the health examination in 1998 were surveyed using self-administered questionnaire. And 230 data were available and analyzed. The results of this study were summarized as follows: Among variables of predisposing factors, knowledge for disease, confidence about periodic health examination program in health insurance, and the attitude toward medical utilization in the usual showed significant relations with the medical utilization. Other variables were not related with the medical utilization. Variables of enabling factors did not show significant relations with the medical utilization. Recognition of family members for detected disease had significant relations with the medical utilization. Among variables of need factors, absence caused by detected disease was significantly related with the medical utilization. The number of non-occupational diseases detected, but untreated people were 75(32.6%) of total subjects, mainly because detected diseases seemed insignificant to them. With multiple logistic regression analysis, the significant variables having an effect on the medical facilities utilization were 'knowledge for disease', 'attitude toward medical utilization in the usual', 'recognition of family members for detected disease' and 'experience of absence caused by detected disease'. On considerations of above findings, counselling for detected disease and its treatment, health education for individuals and program for family support promotion are needed for health management of public officials with diseases detected in health examination.

• Korea journal of population studies
• /
• v.34 no.3
• /
• pp.139-155
• /
• 2011

The purpose of this study was to explore the effective policy on secondary childbirth encouragement of married women. The design for this study is a cross-sectional descriptive study. Samples were collected from 322 women who have one child younger than 6 years old. The instruments for this study was a questionnaire consisted of items about policy on childbirth encouragement. The data were analyzed by chi-square test, Mann-whitney's U test, and logistic regression with SPSS 17.0 program. The results showed that the employed women's intention of having a second child was affected by the age, working day per week, and policy on childbirth. But, in case of unemployed women, both of demographic and policy variables were not statistically significant. These findings suggest that different policy on support of birth, work-family life balance, and various tax benefits be developed to satisfy the needs of the employed women and the unemployed women.

• Journal of The Korean Society of Clinical Toxicology
• /
• v.7 no.2
• /
• pp.77-82
• /
• 2009

Purpose: In highly doses, endosulfan lowers the seizure threshold and elicits central nervous system stimulation, which can result in seizures, respiratory failure, and death. Management of seizure control is essential for survival and prognosis of intoxicated patients. This study assessed whether seizure time was an independent predictor mortality in patients with endosulfan poisoning. Methods: This retrospective study enrolled patients with endosulfan poisoning presenting to Masan Samsung Hospital and Gyeongsang National University Hospital from January 2003 to December 2008. The data were collected from clinical records and laboratory files. Using a multivariate logistic analysis, data on the total population was retrospectively analyzed for association with mortality. Results: Of the 24 patients with endosulfan poisoning, nineteen (79.1%) experienced seizure. The patients in the seizure group showed significantly lower Glasgow coma scale score, base excess, bicarbonate, and significant existence of mechanical ventilation, as compared to the non seizure group (n=5). Seizure, Glasgow coma scale score, systolic blood pressure, bicarbonate level, need for respiratory support, pulse rate, respiratory rate, pH, base excess, and seizure time were associated with mortality. The fatality rate of endosulfan poisoning was 54.1% with higher mortality among patients experiencing. Longer seizure time was associated with higher mortality. Conclusion: Seizure time can be a significant independent predictor of mortality in patients with acute endosulfan poisoning. Physicians should aggressively treat for seizure control in patients with acute endosulfan poisoning.

• Journal of the Korean Dietetic Association
• /
• v.19 no.1
• /
• pp.14-24
• /
• 2013

The purpose of this study was to estimate the prevalence of hypertension, and to assess the risk factors associated with hypertension in elderly (over 65 years old) Koreans, using data from the 4th Korean National Health & Nutrition Examination Survey (the 4th KNHANES), 2007~2009. A total of 1,887 participants were analyzed for assessing the risk factors after excluding those who took hypertension medicines or underwent diet therapy (hypertension perceivers). On the other hand, in analyzing prevalence, a total of 3,526 people were analyzed, including hypertension perceivers. Hypertension was more frequently found in female (64.1%) and urban (62.0%) groups compared to male (55.6%) and rural (57.4%) groups. The mean age of the hypertensive group was significantly higher than that of normal group. The mean values of total cholesterol levels were also significantly higher in the hypertensive group, while body weight, waist circumference, fasting blood glucose, hemoglobin, hematocrit, LDL-cholesterol, and HDL-cholesterol were not. Multiple logistic regression showed that smoking was significantly related to the prevalence of hypertension, but alcohol drinking, stress, obesity, hypercholesterolemia, hypertriglyceridemia, diabetes, anemia, and nutrient intakes were not. The results of this study does not support clear relations of hypertension with chronic diseases including obesity, hyperlipidemia, anemia, and diabetes as well as nutrient intakes among a Korean elderly population. A prospective long-term research study is needed to establish the effects of these factors on hypertension.

• Journal of Information Processing Systems
• /
• v.15 no.6
• /
• pp.1306-1325
• /
• 2019

As the world is moving towards digitization, data is generated from various sources at a faster rate. It is getting humungous and is termed as big data. The financial sector is one domain which needs to leverage the big data being generated to identify financial risks, fraudulent activities, and so on. The design of predictive models for such financial big data is imperative for maintaining the health of the country's economics. Financial data has many features such as transaction history, repayment data, purchase data, investment data, and so on. The main problem in predictive algorithm is finding the right subset of representative features from which the predictive model can be constructed for a particular task. This paper proposes a correlation-based method using submodular optimization for selecting the optimum number of features and thereby, reducing the dimensions of the data for faster and better prediction. The important proposition is that the optimal feature subset should contain features having high correlation with the class label, but should not correlate with each other in the subset. Experiments are conducted to understand the effect of the various subsets on different classification algorithms for loan data. The IBM Bluemix BigData platform is used for experimentation along with the Spark notebook. The results indicate that the proposed approach achieves considerable accuracy with optimal subsets in significantly less execution time. The algorithm is also compared with the existing feature selection and extraction algorithms.

• Journal of Korean Public Health Nursing
• /
• v.13 no.1
• /
• pp.25-40
• /
• 1999

This study was to identify various personal characteristics between those having a helper and those not having a helper for ADL and IADL performance in noninstitutionalized elders. Secondary analysis was done with the data collected in 1994 national survey of aged householders. Of the elders with ADL limitations $47.8\%$ had no helper. while $26.2\%$ of those with IADL limitations had no helper. Those elders with a helper experienced more limitations in ADL and in IADL than those with no helper. Most of the helpers were family members. whereas very few had a formal helper. Age period and subjective economic status were significantly different between the groups with a helper and with no helper for both ADL and IADL. Compared to the elders with a helper. those elders with no helper had fewer children living together. whereas they had more grandchildren. They also had more contacts with friends or neighbors. Health status. in general. was worse in those elders with a helper than those with no helper. Logistic regression analysis revealed that ADL and IADL limitations had largest influence on having an ADL helper. The odds ratios for low levels of subjective economic status were very high. As for having an IADL helper. IADL score was the most important predictor. In particular. ADL limitations had a negative effect on having an IADL helper. This finding may indicate the urgency of ADL needs. which leads to the lack of an IADL helper when both ADL and IADL limitations are present. Considering the nature of ADL. those disabled elders with low economic status need public support for their daily activities and for home visiting nursing services. There should be studies of examining the adequacy of help in satisfying basic needs for performing ADL and IADL in elders.