• Korean Journal of Psychosomatic Medicine
• /
• v.30 no.2
• /
• pp.119-126
• /
• 2022

Objectives : The purpose of this study was to examine the clinical utility of the internalization and externalization higher-order scales of the Minnesota Multiphasic Personality Inventory-Adolescent Restructured Form (MMPI-A-RF), compared with those scales of the Korean Child Behavior Checklist (K-CBCL). Methods : 43 adolescents with internalizing disorders and 44 adolescents with externalizing disorders and their parents were administered the MMPI-A-RF and K-CBCL each. To verify the difference between the internalization and externalization scales of the MMPI-A-RF and K-CBCL for each group, independent-sample t test was performed. To compare the agreement between the MMPI-A-RF and K-CBCL, correlation analysis was also conducted. Lastly, to identify which scales significantly best predict each of the internalizing and externalizing disorder, logistic regression analysis was conducted. Results : Internalization scales of the MMPI-A-RF and K-CBCL were significantly higher in the internalizing disorder group, and the externalization scales were significantly higher in the externalizing disorder group. The positive correlation was significant only for internalization problems between the two evaluation measures in both groups (each r=0.360, p<0.05, r=0.572, p<0.05). In addition, the scales significantly predicted internalizing and externalizing disorders were the internalization and externalization scales of the MMPI-A-RF, followed by the externalization scale of the K-CBCL (R²=0.407, p<0.05). Conclusions : The internalization and externalization higher-order scales of the MMPI-A-RF were found to reliably reflect the characteristics of each disorder in adolescents and be useful evaluative scales to differentiate disorders. Moreover, if adolescents show externalization problems, additional information from the K-CBCL can be more useful to differentiate disorders.

• Asia-Pacific Journal of Business Venturing and Entrepreneurship
• /
• v.18 no.1
• /
• pp.253-270
• /
• 2023

Research on investment determinants of accelerators, which are attracting attention by greatly improving the survival rate of startups by providing professional incubation and investment to startups at the same time, is gradually expanding. However, previous studies do not have a theoretical basis in developing investment determinants in the early stages, and they use factors of angel investors or venture capital, which are similar investors, and are still in the stage of analyzing importance and priority through empirical research. Therefore, this study verified for the first time in Korea the discrimination and effectiveness of investment determinants using accelerator investment determinants developed based on the business model innovation framework in previous studies. To this end, we first set the criteria for success and failure of startup investment based on scale-up theory and conducted a survey of 22 investment experts from 14 accelerators in Korea, and secured valid data on a total of 97 startups, including 52 successful scale-up startups and 45 failed scale-up startups, were obtained and an independent sample t-test was conducted to verify the mean difference between these two groups by accelerator investment determinants. As a result of the analysis, it was confirmed that the investment determinants of accelerators based on business model innovation framework have considerable discrimination in finding successful startups and making investment decisions. In addition, as a result of analyzing manufacturing-related startups and service-related startups considering the characteristics of innovation by industry, manufacturing-related startups differed in business model, strategy, and dynamic capability factors, while service-related startups differed in dynamic capabilities. This study has great academic implications in that it verified the practical effectiveness of accelerator investment determinants derived based on business model innovation framework for the first time in Korea, and it has high practical value in that it can make effective investments by providing theoretical grounds and detailed information for investment decisions.

• The korean journal of orthodontics
• /
• v.25 no.4
• /
• pp.465-485
• /
• 1995

The purpose of this study is to locate the proper position of the lower occlusal plane according to individual skeletal pattern. Cephalometric films of 234 subjects of the control group, 358 of the pretreatment group and 358 of the treated group were analyzed to study proper relationships between vertical dimension ratio(VDR) and lower occlusomandibular plane angle(LOM). The control group was divided into two subgroups by the age. The first subgroup consisted of 113 subjects of the age 14 years and under and with the mean age of 10.82 years. The other subgroup consisted of 113 subjects of the age 18 years and above with the mean age of 23.76 years. The pretreatment group was divided into three subgroups by the age. The first subgroup consisted of 274 subjects of the age 14 years and under with the mean age of 11.36 years. The second subgroup consisted of 54 subjects of the age 14 through 18 years with the mean age of 15.4 years. The last subgroup consisted of 30 subjects of the age 18 years and above with the mean age of 21.35 years. The treated group was also divided into three subgroups by the age. The first subgroup consisted of 145 subjects of the age 14 years and under with the mean age of 12.91 years. The second subgroup consisted of 166 subjects of the age 14 through 18 years with the mean age of 15.64 years. The last subgroup consisted of 47 subjects of the age 18 years and above with the mean age of 21.61 years. Cephalometric films of the sample were traced. Measurements were made to a hundredth using a program specifically prepared for this study, and the results were entered into a 486DX PC. Means and Standard deviations of all the veriables were calculated for each group. Correlation coefficients between pertinent variables were calculated. Significance tests on those coefficients, one-way ANOVA and t-tests between variables or groups were performed. On the basis of the results studied above, certain subjects were selected from the control and the treated groups to locate the proper position of the occlusal plane, and designated as the optimal occluaion group. The subjects of this optimal occlusion group had 1-3 mm overbite, 1-3 mm of overjet and less than 1.75 mm of curve of Spee. A total subjects of 187 in this group consisted 104 treated subjects and 83 control group. Regression analysis was carried out between VDR and LOM, and regression equations were tabulated for this optimal occlusion group. The results were as follows : 1. Highly significant correlations were observed between various variables useful for identifying vertical component of skeletal frame, but any one particular variable did not accurately indicate the magnitude of anterior vertical overbite. 2. Of the variables useful identifying vertical component of skeletal frame, The VDR showed the highest correlation to the LOM. 3. Of the total sample, 80 percent had overbite within the normal range, irrespective of VDR. 4. The optimal occlusion group was divided into 9 subgroups by the age and the anteroposterior skeletal pattern, and correlation coefficient and determination coefficient between VDR and LOM of each group were calculated. Correlation coefficients and determination coefficients were found to be significantly high in all groups. 5. Regression equation was induced for each of the optimal occlusion group to find proper LOM according to the VDR. 6. It was found that the mean value of the cant of occlusal plane itself is not enough for a diagnosis and a treatment plan. Rather, It is very important to locate the proper occlusal plane for an Individual skeletal pattern.

• Journal of Korean Academy of Nursing
• /
• v.2 no.1
• /
• pp.3-33
• /
• 1971

Nursing of today has as one of its objectives the solving of problems related to human needs arising from the demands of a rapidly changing society. This nursing objective, I believe, can he attained by the appropriate application of scientific principles in the giving of comprehensive nursing care. Comprehensive nursing care may be defined as nursing care which meets all of the patient's needs. the needs of patients are said to fall into five broad categories: physical needs, psychological needs, environmental needs, socio-economic needs, and teaching needs. Most people who become ill have adjustment problems related to their new situation. Because patient teaching is one of the most important functions of professional nursing, the success of this teaching may be used as a gauge for evaluating comprehensive nursing care. This represents a challenge foe the future. A questionnaire consisting of 67 items was distributed to 200 professional nurses working ill direct patient care at Yonsei University Medical Center in Seoul, Korea. 160 (80,0%) nurses of the total sample returned completed questionnaires 81 (50.6%) nurses were graduates of 3 fear diploma courser 79 (49.4%) nurses were graduates of 4 year collegiate nursing schools in Korea 141 (88,1%) nurses had under 5 years of clinical experience in a medical center, while 19 (11.9%) nurses had more than 5years of clinical experience. Three hypotheses were tested: 1. “Nurses had high levels of concept and knowledge toward patient teaching”-This was demonstrated by the use of a statistical method, the mean average. 2. “Nurses graduating from collegiate programs and diploma school programs of nursing show differences in concepts and knowledge toward patient teaching”-This was demonstrated by a statistical method, the mean average, although the results showed little difference between the two groups. 3. “Nurses having different amounts of clinical experience showed differences in concepts and knowledge toward patient teaching”-This was demonstrated by the use of a statistical method, the mean average. 2. “Nurses graduating from collegiate programs and diploma school programs of nursing show differences in concepts and knowledge toward patient teaching”-This was demonstrated by a statistical method, the mean average, although the results showed little difference between the two groups. 3. “Nurses having different amounts of clinical experience showed differences in concepts and knowledge toward patient teaching”-This was demonstrated by the use of the T-test. Conclusions of this study are as follow: Before attempting the explanation, of the results, the questionnaire will he explained. The questionnaire contained 67 questions divided into 9 sections. These sections were: concept, content, time, prior preparation, method, purpose, condition, evaluation, and recommendations for patient teaching. 1. The nurse's concept of patient teaching: Most of the nurses had high levels of concepts and knowledge toward patient teaching. Though nursing service was task-centered at the turn of the century, the emphasis today is put on patient-centered nursing. But we find some of the nurses (39.4%) still are task-centered. After, patient teaching, only a few of the nurses (14.4%) checked this as “normal teaching.”It seems therefore that patient teaching is often done unconsciously. Accordingly it would he desirable to have correct concepts and knowledge of teaching taught in schools of nursing. 2. Contents of patient teaching: Most nurses (97.5%) had good information about content of patient teaching. They teach their patients during admission about their diseases, tests, treatments, and before discharge give nurses instruction about simple nursing care, personal hygiene, special diets, rest and sleep, elimination etc. 3. Time of patient teaching: Teaching can be accomplished even if there is no time set aside specifically for it. -a large part of the nurse's teaching can be done while she is giving nursing care. If she believes she has to wait for time free from other activities, she may miss many teaching opportunities. But generally proper time for patient teaching is in the midmorning or midafternoon since one and a half or two hours required. Nurses meet their patients in all stages of health: often tile patient is in a condition in which learning is impossible-pain, mental confusion, debilitation, loss of sensory perception, fear and anxiety-any of these conditions may preclude the possibility of successful teaching. 4. Prior preparation for patient teaching: The teaching aids, nurses use are charts (53.1%), periodicals (23.8%), and books (7.0%) Some of the respondents (28.1%) reported that they had had good preparation for the teaching which they were doing, others (27.5%) reported adequate preparation, and others (43.8%) reported that their preparation for teaching was inadequate. If nurses have advance preparation for normal teaching and are aware of their objectives in teaching patients, they can do effective teaching. 5. Method of patient teaching: The methods of individual patient teaching, the nurses in this study used, were conversation (55.6%) and individual discussion (19.2%) . And the methods of group patient teaching they used were demonstration (42.3%) and lecture (26.2%) They should also he prepared to use pamphlet and simple audio-visual aids for their teaching. 6. Purposes of patient teaching: The purposes of patient teaching is to help the patient recover completely, but the majority of the respondents (40.6%) don't know this. So it is necessary for them to understand correctly the purpose of patient teaching and nursing care. 7. Condition of patient teaching: The majority of respondents (75.0%) reported there were some troubles in teaching uncooperative patients. It would seem that the nurse's leaching would be improved if, in her preparation, she was given a better understanding of the patient and communication skills. The majority of respondents in the total group, felt teaching is their responsibility and they should teach their patient's family as well as the patient. The place for teaching is most often at the patient's bedside (95.6%) but the conference room (3.1%) is also used. It is important that privacy be provided in learning situations with involve personal matters. 8. Evaluation of patient teaching: The majority of respondents (76.3%,) felt leaching is a highly systematic and organized function requiring special preparation in a college or university, they have the idea that teaching is a continuous and ever-present activity of all people throughout their lives. The suggestion mentioned the most frequently for improving preparation was a course in patient teaching included in the basic nursing program. 9. Recommendations: 1) It is recommended, that in clinical nursing, patient teaching be emphasized. 2) It is recommended, that insertive education the concepts and purposes of patient teaching he renewed for all nurses. In addition to this new knowledge, methods and materials which can be applied to patient teaching should be given also. 3) It is recommended, in group patient teaching, we try to embark on team teaching.

• Journal of Intelligence and Information Systems
• /
• v.18 no.2
• /
• pp.29-45
• /
• 2012

Bond rating is regarded as an important event for measuring financial risk of companies and for determining the investment returns of investors. As a result, it has been a popular research topic for researchers to predict companies' credit ratings by applying statistical and machine learning techniques. The statistical techniques, including multiple regression, multiple discriminant analysis (MDA), logistic models (LOGIT), and probit analysis, have been traditionally used in bond rating. However, one major drawback is that it should be based on strict assumptions. Such strict assumptions include linearity, normality, independence among predictor variables and pre-existing functional forms relating the criterion variablesand the predictor variables. Those strict assumptions of traditional statistics have limited their application to the real world. Machine learning techniques also used in bond rating prediction models include decision trees (DT), neural networks (NN), and Support Vector Machine (SVM). Especially, SVM is recognized as a new and promising classification and regression analysis method. SVM learns a separating hyperplane that can maximize the margin between two categories. SVM is simple enough to be analyzed mathematical, and leads to high performance in practical applications. SVM implements the structuralrisk minimization principle and searches to minimize an upper bound of the generalization error. In addition, the solution of SVM may be a global optimum and thus, overfitting is unlikely to occur with SVM. In addition, SVM does not require too many data sample for training since it builds prediction models by only using some representative sample near the boundaries called support vectors. A number of experimental researches have indicated that SVM has been successfully applied in a variety of pattern recognition fields. However, there are three major drawbacks that can be potential causes for degrading SVM's performance. First, SVM is originally proposed for solving binary-class classification problems. Methods for combining SVMs for multi-class classification such as One-Against-One, One-Against-All have been proposed, but they do not improve the performance in multi-class classification problem as much as SVM for binary-class classification. Second, approximation algorithms (e.g. decomposition methods, sequential minimal optimization algorithm) could be used for effective multi-class computation to reduce computation time, but it could deteriorate classification performance. Third, the difficulty in multi-class prediction problems is in data imbalance problem that can occur when the number of instances in one class greatly outnumbers the number of instances in the other class. Such data sets often cause a default classifier to be built due to skewed boundary and thus the reduction in the classification accuracy of such a classifier. SVM ensemble learning is one of machine learning methods to cope with the above drawbacks. Ensemble learning is a method for improving the performance of classification and prediction algorithms. AdaBoost is one of the widely used ensemble learning techniques. It constructs a composite classifier by sequentially training classifiers while increasing weight on the misclassified observations through iterations. The observations that are incorrectly predicted by previous classifiers are chosen more often than examples that are correctly predicted. Thus Boosting attempts to produce new classifiers that are better able to predict examples for which the current ensemble's performance is poor. In this way, it can reinforce the training of the misclassified observations of the minority class. This paper proposes a multiclass Geometric Mean-based Boosting (MGM-Boost) to resolve multiclass prediction problem. Since MGM-Boost introduces the notion of geometric mean into AdaBoost, it can perform learning process considering the geometric mean-based accuracy and errors of multiclass. This study applies MGM-Boost to the real-world bond rating case for Korean companies to examine the feasibility of MGM-Boost. 10-fold cross validations for threetimes with different random seeds are performed in order to ensure that the comparison among three different classifiers does not happen by chance. For each of 10-fold cross validation, the entire data set is first partitioned into tenequal-sized sets, and then each set is in turn used as the test set while the classifier trains on the other nine sets. That is, cross-validated folds have been tested independently of each algorithm. Through these steps, we have obtained the results for classifiers on each of the 30 experiments. In the comparison of arithmetic mean-based prediction accuracy between individual classifiers, MGM-Boost (52.95%) shows higher prediction accuracy than both AdaBoost (51.69%) and SVM (49.47%). MGM-Boost (28.12%) also shows the higher prediction accuracy than AdaBoost (24.65%) and SVM (15.42%)in terms of geometric mean-based prediction accuracy. T-test is used to examine whether the performance of each classifiers for 30 folds is significantly different. The results indicate that performance of MGM-Boost is significantly different from AdaBoost and SVM classifiers at 1% level. These results mean that MGM-Boost can provide robust and stable solutions to multi-classproblems such as bond rating.

• Journal of Korean Public Health Nursing
• /
• v.15 no.2
• /
• pp.301-313
• /
• 2001

This study was to explore the relationship between nurses' personality type and their job satisfaction level. The subjects of the study were the 364 nurses who were serving at the general surgical department, psychiatric department, pediatric department and emergency room. The hospitals for which they were working were four university hospitals in Daegu and Busan, and six other university hospitals in Seoul were selected as sample hospitals. For data collection, questionnaires were self-reported to the subjects, under their agreement, from May to June, 2000. The following two instruments were used in the study: One was MBTI Test whose Korean version designed and verified to reliability and validity by Kim Jung Taek and Shim Hae Sook(1990): the other was the Nursing Job Satisfaction Instrument, 'The Index of Work Satisfaction' developed by Slavitt, et al.., (1978). The data were analyzed by frequency, t-test, one-way variance analysis, scheffe's post hoc contrast, and Pearson correlation coefficient with SPSS Win 10.0 program. The results this study were as follows ; 1. As a result of testing nurse's personality type by function at each career, it's found that the most personality type was expressed by the ST-type$(37.3\%)$ at 1-5 years, the SF-type$(29.4\%)$ at 6-10 years, the ST-type$(43.3\%)$ at more than 11 years. 2. As a result of testing the disparity in job satisfaction according to career, it appeared that the 6-10 years group showed higher job satisfaction, followed by the 6-10 years$(119.6\%),\;1-5 years(118.6\%)\;and\;more\;than\; 11 years(117.7\%)$ groups. 3. As a result of testing job satisfaction level according to nurse's personality type by function at each career, it's found that the most job satisfaction was expressed by the SF-type(F=8.50, p=0.00l) at 1-5 years. the ST-type(F=30.61, p=0.001) at 6-10 years. the SF-type(F=4.98, p=0.003) at more than 11 years groups. 4. As a result of testing a significant correlation between nurse's personality type by function and job satisfaction level. the SF(r=0.279, p=0.001) and ST(r=0.222, p=0.001) types showed significant positive correlation. and the NF(r=-0.201, p=0.001) and NT(r=-0.402, p=0.001) types revealed significant negative correlation. The nursing management is likely to be done more effectively, if managers in charge of nursing administration carefully reflect nurses' personal opinions in posting and personnel management, keeping in mind that there is a specific personality type that serves to raise job satisfaction at a specific career.

• Women's Health Nursing
• /
• v.5 no.1
• /
• pp.111-132
• /
• 1999

This correlational descriptive study sought to define the relationship between the experience of Sanhujori, Korean tradition non-professional postpartal care after delivery and abortion and present physical symptom distress of arthritis female who visited to outpatient clinic of rheumatic internal medicine at three hospital located in Seoul, Pusan, Chongju, Korea. Data from a convenience sample of 98 women who orally agreed to be respondent were collected from September 1, 1998 to October 31, 1998 for two months by way of interview with semi-structured questionnaire. Data analysis consisted of frequency, percentage, mean, S. D., Pearson Correlation Coefficient, t-test, ANOVA and Scheffe test as a post hoc by SPSS. The results of the study were as follows ; Mean age of participants as 52.8 years and mean number of children 3.3. Mean frequency of child birth was 3.1 times per woman, 67.4% of respondents had menopause, 57.0% did not have Sunhujori after abortion. The health status implies the subjective health status women perceived, which came from the three points of view of the present, comparative with other of same age and changed after delivery. The respondents of 76.1% perceived them as unhealthy and the main sites of physical symptom distress were upper & lower extremities including knee and hand 34.8%, shoulder 26.5%, waist 22.4%. Women perceived the etiology of the arthritis as 'did Sanhujori wrongly' 36.7%, 'aging process' 24.5%, 'stress' 16.3%, 'overwork' 15.3%, 'Immunocompromize' & 'physical constitution' 7.1% respectively, 'character' 3.1%, 'genetic' 2.0%, 'malnutrition' 1.0%. The mean period of Sanhujori after delivery was 20.4 at the first child and 18.1 at the second child. The higher frequency of child birth, the shorter period of Sanhujori. For the subjective evaluation of whether the women did Sanhujori well or not, the rate of 'did Sanhujori well' was the highest rank in the first child and the rate of 'did not particularly Sanhujori well' was the highest rank in the last child. There was a significant positive correlation between physical function disability and rheumatoid arthritis symptom at the level of 1% of significance statistically(r=.406). And a positive correlation between physical function disability and Sanhubyung symptom at the level of 5% of significance statistically(r=.224). There was a significant positive correlation among rheumatoid arthritis symptom, Sanhubyung symptom and menopause symptom at the level of 1%-5% of significance statistically. Most of all, the correlation between Sanhubyung symptom and menopause symptom was the higher than others. There was a negative correlation between the present physical symptom distress and experience of Sanhujori(r=.-130), however it was not significant statistically. However, there was a positive correlation between subjective health status and experience of Sanhujori at the level of 1% of significance statistically(r=.328). In conclusion, this finding reconfirmed the positive relationship between the perceived health status and experience of Sanhujori after delivery among women of arthritis. It provides a challenge to the professional care givers to study further on the effects of Sanhuiori after abortion or delivery on the physical symptom distress from the variouis aspects through the cross-sectional and longitudinal research. The strategy for the development of the appropriate intervention for primary prevention of sequele after childbirth and quality of care for desirable health outcomes for postpartal women with considering deeply on the relationship between women's health and postapartal care.

• Journal of Periodontal and Implant Science
• /
• v.36 no.2
• /
• pp.555-565
• /
• 2006

The purpose of the present study is to evaluate the biological stability of the zirconia/alumina composite abutment by histologic and radiographic examination in clinical cases. 17 partially edentulous patients (5 men and 12 women, mean age 47) were treated with 37 implants. The implants were placed following the standard two-stage protocol. After a healing period of 3 to 6 months, zirconia/alumina composite abutments were connected. All radiographs were taken using paralleling technique with individually fabricated impression bite block, following insertion of the prosthesis and at the 3-, 6-, 12 month re-examinations. After processing the obtained images, the osseous level was calculated using the digital image in the mesial and distal aspect in each implant. An ANOVA and t-test were used to test for difference between the baseline and 3-, 6-, 12 months re-examinations, and for difference between maxilla and mandible. Differences at P <0.05 were considered statistically significant. For histologic examination, sample was obtained from the palatal gingiva which implant functioned for 12 months. Sections were examined under a light microscope under various magnifications. Clinically, no abutment fracture or crack as well as periimplantitis was observed during the period of study. The mean bone level reduction(${\pm}standard$ deviation) was 0.34 rom(${\pm}\;0.26$) at 3-months, 0.4 2mm(${\pm}\;0.30$) at 6-months, 0.62 mm(${\pm}\;0.28$) at 12-months respectively. No statistically significant difference was found between baseline and 3-, 6-, 12-months re-examinations (p > 0.05). The mean bone level reduction in maxilla was 0.33(${\pm}0.25$) at 3-months, 0.36(${\pm}0.33$) at 6-months, 0.56(${\pm}0.26$) at 12-months. And the mean bone level reduction in mandible was 0.35(${\pm}0.27$) at 3-months, 0,49(${\pm}0.27$) at 6-months, 0.68(${\pm}0.30$) at 12-months. No statistical difference in bone level reduction between implants placed in the maxilla and mandible. Histologically, the height of the junctional epithelium was about 2.09 mm. And the width was about 0.51 mm. Scattered fibroblasts and inflammatory cells, and dense collagen network with few vascular structures characterized the portion of connective tissue. The inflammatory cell infiltration was observed just beneath the apical end of junctional epithelium and the area of direct in contact with zirconia/alumina abutment. These results suggest the zirconia/alumina composite abutment can be used in variable intraoral condition, in posterior segment as well as anterior segment without adverse effects.

• Journal of Nutrition and Health
• /
• v.56 no.3
• /
• pp.300-314
• /
• 2023

Purpose: This study sought to predict the reasons for skipping breakfast by adolescents aged 13-18 years using the 7^th Korea National Health and Nutrition Examination Survey (KNHANES). Methods: The participants included 1,024 adolescents. The data were analyzed using a complex-sample t-test, the Rao Scott χ²-test, and the classification and regression tree (CART) algorithm for decision tree analysis with SPSS v. 27.0. The participants were divided into two groups, one regularly eating breakfast and the other skipping it. Results: A total of 579 and 445 study participants were found to be breakfast consumers and breakfast skippers respectively. Breakfast consumers were significantly younger than those who skipped breakfast. In addition, breakfast consumers had a significantly higher frequency of eating dinner, had been taught about nutrition, and had a lower frequency of eating out. The breakfast skippers did so to lose weight. Children who skipped breakfast consumed less energy, carbohydrates, proteins, fats, fiber, cholesterol, vitamin C, vitamin A, calcium, vitamin B1, vitamin B2, phosphorus, sodium, iron, potassium, and niacin than those who consumed breakfast. The best predictor of skipping breakfast was identifying adolescents who sought to control their weight by not eating meals. Other participants who had low and middle-low household incomes, ate dinner 3-4 times a week, were more than 14.5 years old, and ate out once a day showed a higher frequency of skipping breakfast. Conclusion: Based on these results, nutrition education targeted at losing weight correctly and emphasizing the importance of breakfast, especially for adolescents, is required. Moreover, nutrition educators should consider designing and implementing specific action plans to encourage adolescents to improve their breakfast-eating practices by also eating dinner regularly and reducing eating out.

• Journal of Intelligence and Information Systems
• /
• v.22 no.1
• /
• pp.139-157
• /
• 2016

Bankruptcy involves considerable costs, so it can have significant effects on a country's economy. Thus, bankruptcy prediction is an important issue. Over the past several decades, many researchers have addressed topics associated with bankruptcy prediction. Early research on bankruptcy prediction employed conventional statistical methods such as univariate analysis, discriminant analysis, multiple regression, and logistic regression. Later on, many studies began utilizing artificial intelligence techniques such as inductive learning, neural networks, and case-based reasoning. Currently, ensemble models are being utilized to enhance the accuracy of bankruptcy prediction. Ensemble classification involves combining multiple classifiers to obtain more accurate predictions than those obtained using individual models. Ensemble learning techniques are known to be very useful for improving the generalization ability of the classifier. Base classifiers in the ensemble must be as accurate and diverse as possible in order to enhance the generalization ability of an ensemble model. Commonly used methods for constructing ensemble classifiers include bagging, boosting, and random subspace. The random subspace method selects a random feature subset for each classifier from the original feature space to diversify the base classifiers of an ensemble. Each ensemble member is trained by a randomly chosen feature subspace from the original feature set, and predictions from each ensemble member are combined by an aggregation method. The k-nearest neighbors (KNN) classifier is robust with respect to variations in the dataset but is very sensitive to changes in the feature space. For this reason, KNN is a good classifier for the random subspace method. The KNN random subspace ensemble model has been shown to be very effective for improving an individual KNN model. The k parameter of KNN base classifiers and selected feature subsets for base classifiers play an important role in determining the performance of the KNN ensemble model. However, few studies have focused on optimizing the k parameter and feature subsets of base classifiers in the ensemble. This study proposed a new ensemble method that improves upon the performance KNN ensemble model by optimizing both k parameters and feature subsets of base classifiers. A genetic algorithm was used to optimize the KNN ensemble model and improve the prediction accuracy of the ensemble model. The proposed model was applied to a bankruptcy prediction problem by using a real dataset from Korean companies. The research data included 1800 externally non-audited firms that filed for bankruptcy (900 cases) or non-bankruptcy (900 cases). Initially, the dataset consisted of 134 financial ratios. Prior to the experiments, 75 financial ratios were selected based on an independent sample t-test of each financial ratio as an input variable and bankruptcy or non-bankruptcy as an output variable. Of these, 24 financial ratios were selected by using a logistic regression backward feature selection method. The complete dataset was separated into two parts: training and validation. The training dataset was further divided into two portions: one for the training model and the other to avoid overfitting. The prediction accuracy against this dataset was used to determine the fitness value in order to avoid overfitting. The validation dataset was used to evaluate the effectiveness of the final model. A 10-fold cross-validation was implemented to compare the performances of the proposed model and other models. To evaluate the effectiveness of the proposed model, the classification accuracy of the proposed model was compared with that of other models. The Q-statistic values and average classification accuracies of base classifiers were investigated. The experimental results showed that the proposed model outperformed other models, such as the single model and random subspace ensemble model.