Classification of histogram-valued data with support histogram machines.

The current large amounts of data and advanced technologies have produced new types of complex data, such as histogram-valued data. The paper focuses on classification problems when predictors are observed as or aggregated into histograms. Because conventional classification methods take vectors as input, a natural approach converts histograms into vector-valued data using summary values, such as the mean or median. However, this approach forgoes the distributional information available in histograms. To address this issue, we propose a margin-based classifier called support histogram machine (SHM) for histogram-valued data. We adopt the support vector machine framework and the Wasserstein-Kantorovich metric to measure distances between histograms. The proposed optimization problem is solved by a dual approach. We then test the proposed SHM via simulated and real examples and demonstrate its superior performance to summary-value-based methods.

Convex clustering analysis for histogram-valued data.

In recent years, there has been increased interest in symbolic data analysis, including for exploratory analysis, supervised and unsupervised learning, time series analysis, etc. Traditional statistical approaches that are designed to analyze single-valued data are not suitable because they cannot incorporate the additional information on data structure available in symbolic data, and thus new techniques have been proposed for symbolic data to bridge this gap. In this article, we develop a regularized convex clustering approach for grouping histogram-valued data. The convex clustering is a relaxation of hierarchical clustering methods, where prototypes are grouped by having exactly the same value in each group via penalization of parameters. We apply two different distance metrics to measure (dis)similarity between histograms. Various numerical examples confirm that the proposed method shows better performance than other competitors.

Evaluating the prevalence, awareness, and control of hypertension, diabetes, and dyslipidemia in Korea using the NHIS-NSC database: A cross-sectional analysis.

This study aimed to evaluate the prevalence, awareness, and control of hypertension, diabetes, and dyslipidemia in Korea according to sex and age, and to identify factors that were associated with the prevalence, awareness, and control of these cardiovascular risk factors among middle-aged Koreans.A cross-sectional analysis was conducted using 2013 data from the National Health Insurance Service-National Sample Cohort (NHIS-NSC) database. A total of 136,755 individuals were considered eligible based on no history of stroke or heart disease, age of 40 to 64 years at the examination, and completion of health examinations. Data regarding demographic factors and cardiovascular risk factors had been obtained using questionnaires, physical examinations, and blood tests.Hypertension was the most prevalent condition among middle-aged Koreans (25.76%), followed by dyslipidemia (16.58%) and diabetes (10.22%). Among these three conditions, dyslipidemia was associated with the lowest awareness rates (24.14%). Hypertension and diabetes were more prevalent among men than among women, although the prevalence of dyslipidemia was higher among women > 50 years old. The prevalence of hypertension and diabetes increased with advancing age among men and women, although the prevalence of dyslipidemia decreased with age among men (highest prevalence: 50.59% among men aged between 40 and 44 years). Among Korean men (relative to women), the odds ratios were 1.50 for hypertension, 1.96 for diabetes, and 0.82 for dyslipidemia. Age, central obesity, and body mass index were significantly associated with the development of all 3 conditions.There are sex- and age-specific patterns in the prevalence, awareness, and control of hypertension, diabetes, and dyslipidemia in Korea. Dyslipidemia is becoming increasingly common although most middle-aged Koreans are not aware of this condition leading to low control rate. These results may help identify population subgroups at a high risk of hypertension, diabetes, and dyslipidemia, and may guide the management of cardiovascular disease in Korea.

Workplace Bullying, Job Stress, Intent to Leave, and Nurses' Perceptions of Patient Safety in South Korean Hospitals.

BACKGROUND: Negative work environments influence the ability of nurses to provide optimal patient care in a safe environment. AIM: The purpose of the study was to test a model linking workplace bullying (WPB) and lateral violence (LV) with job stress, intent to leave, and, subsequently, nurse-assessed patient adverse outcomes (safety issues). DESIGN: This descriptive-correlational study examined the relationships between study variables and used a structural equation model to test the validity of the proposed theoretical framework. METHODS: A convenience sample of 508 clinical nurses working in eight general hospitals in Daejeon, South Korea, completed a questionnaire on measures of WPB, LV, job stress, intent to leave, and nurse-assessed patient safety. Analysis of moment structures was used to estimate a set of three models with competing measurement structures for WPB and LV and the same structural model. Akaike Information Criterion was used for model selection. RESULTS: Among the three proposed models, the model with complex factor loadings was selected (WPB and LV were both associated with verbal abuse and physical threat). WPB directly and indirectly influenced nurse-assessed patient safety. Job stress directly influenced intent to leave, and intent to leave directly influenced nurse-assessed patient safety. CONCLUSIONS: The results of the study support the proposition that WPB, job stress, and intent to leave may be associated with nurse-perceived adverse outcomes (patient safety issues) in hospitals. Nurse perceptions of WPB were associated with nurse-assessed patient safety outcomes (adverse events) directly and through mediating job stress and intent to leave. LV was not associated with the mediators or nurse-assessed adverse outcomes (safety).

Factors affecting workplace bullying and lateral violence among clinical nurses in Korea: descriptive study.

BACKGROUND & AIM: Workplace bullying and lateral violence are serious issues affecting the work life of hospital nurses. The purpose of this study was to identify the selected individual and institutional characteristics for workplace bullying and lateral violence using a conceptual framework. METHODS: A descriptive survey design was used. A convenience sample of 255 nurses in tertiary hospitals, who had a minimum of 6 months clinical experience, completed the survey. Regression analysis was used to determine factors significantly associated with workplace bullying and lateral violence. The Negative Acts Questionnaire-Revised and the Lateral Violence scale were used to measure workplace bullying and lateral violence. RESULTS: A negative affect, individualism and working in hospital specialty units predicted workplace bullying. Individualism, a negative affect, affiliated hospital and working hours predicted verbal abuse whereas the place of employment was significantly associated with lateral violence. CONCLUSION: The results of this study identified factors that are associated with bullying and violence but did not fully support the conceptual framework. The individual characteristic negative affect was significantly associated with most types of workplace bullying and lateral violence bully whereas the place of employment was an important factor in lateral violence. IMPLICATIONS FOR NURSING MANAGEMENT: Nurse managers need to be aware that both individual and institutional factors may impact levels of workplace bullying and lateral violence in their hospitals and need to prepare specific strategies to address these multiple factors.

Prevalence of putative periodontopathogens in subgingival dental plaques from gingivitis lesions in Korean orthodontic patients.

The objective of this study was to detect and compare the presence of periodontopathogens in the subgingival plaques of gingivitis lesions in adults who wore fixed orthodontic appliances, as opposed to adults who did not wear any orthodontic appliances. Thirty-six individuals participated in this study. Nineteen of these subjects did not wear any orthodontic appliances, and these subjects comprised the control group. The other 17 individuals had been wearing fixed orthodontic appliances for at least 3 months each. After a periodontal examination, we collected subgingival plaque samples from the gingivitis lesions of each patient. Using PCR based on 16S rDNA, we detected the presence of 6 putative periodontopathogenic species, Treponema denticola, Porphyromonas gingivalis, Tannerella forsythia (formerly Bacteroides forsythus), Prevotella nigrescens, Prevotella intermedia, and Actinobacillus actinomycetemcomitans. With regard to the presence of individual periodontopathogens, we found that T. forsythia, T. denticola, and P. nigrescens were significantly more common in the samples obtained from the orthodontic patients than in the samples obtained from the non-orthodontic patient controls. Our results indicate that the local changes associated with the wearing of fixed orthodontic appliances may affect the prevalence of periodontopathogens in subgingival dental plaques.