• Journal of Distribution Science
• /
• v.11 no.3
• /
• pp.5-11
• /
• 2013

Purpose - Going concern is one of fundamental concepts in accounting and auditing and sometimes the assessment of a company's going concern status that is a tough process. Various going concern prediction models' based on statistical and data mining methods help auditors and stakeholders suggested in the previous literature. Research design - This paper employs a data mining approach to prediction of going concern status of Iranian firms listed in Tehran Stock Exchange using Particle Swarm Optimization. To reach this goal, at the first step, we used the stepwise discriminant analysis it is selected the final variables from among of 42 variables and in the second stage; we applied a grid-search technique using 10-fold cross-validation to find out the optimal model. Results - The empirical tests show that the particle swarm optimization (PSO) model reached 99.92% and 99.28% accuracy rates for training and holdout data. Conclusions - The authors conclude that PSO model is applicable for prediction going concern of Iranian listed companies.

• Proceedings of the KOSOMBE Conference
• /
• v.1992 no.05
• /
• pp.155-160
• /
• 1992

Liver function test is one of the most common tests for diagnosis and follow-up of patients and for heal th screening. Automatic interpretation and suggestions on the diagnostic possibilities contribute to shorten the interpretation time of the test results and help to provide qualified health care. Fuzzy logic has been recently introduced and being spread for these purposes. The present study aims at model Ins the foray rule-based laboratory diagnosis system. The fuzzy rule-based laboratory diagnosis system was applied to the diagnosis regarding liver function test. The system was evaluated by comparing with the stepwise multivariate discriminant function analysis, which showed similar results, and the overall accuracy of the fuzzy diagnosis system was about 80%.

• Asia pacific journal of information systems
• /
• v.21 no.2
• /
• pp.43-58
• /
• 2011

Support vector machines (SVMs), a machine learning technique, has been applied to not only binary classification problems such as bankruptcy prediction but also multi-class problems such as corporate credit ratings. However, in general, the performance of SVMs can be easily worse than the best alternative model to SVMs according to the selection of predictors, even though SVMs has the distinguishing feature of successfully classifying and predicting in a lot of dichotomous or multi-class problems. For overcoming the weakness of SVMs, this study has proposed an approach for selecting features for multi-class SVMs that utilize the impurity measures of classification trees. For the selection of the input features, we employed the C4.5 and CART algorithms, including the stepwise method of discriminant analysis, which is a well-known method for selecting features. We have built a multi-class SVMs model for credit rating using the above method and presented experimental results with data regarding S&P 500 companies.

• Korean Journal of Veterinary Research
• /
• v.43 no.1
• /
• pp.165-169
• /
• 2003

The aim of the present study was to analyse grunts of the sows and to extract parameters from the time and frequency signals in nursing behavior. Five crossbred $Landrace{\times}Yorkshire$ sows were used on day 5 or 6 postpartum. The grunts and the behaviors of the five sows were recorded with five digital camcorders. Three parameter groups [Group I: Formant vector alone, Group II: Formant vector+parameters from time signal, Group III: Formant vector+parameters from time signal-parameters eliminated by stepwise discriminant analysis backward (SDAB)] with parameter vectors extracted from single grunts in the maximum grunting rate period were used for individuality of the sows. The parameter groups were compared by a discriminant function analysis. The classification system adopted in the Group II represented the higher discriniation rate than those in other groups (Group I: 63.3%, Group II: 83.0%, Group III: 80.0%). This study demonstrated that formant, intensity, and pitch were available sound parameters for individuality of the sows during nursing behavior.

• Journal of Korean Institute of Industrial Engineers
• /
• v.40 no.1
• /
• pp.8-17
• /
• 2014

In this research, we employed various data mining techniques to build predictive models for win-loss prediction in Korean professional baseball games. The historical data containing information about players and teams was obtained from the official materials that are provided by the KBO website. Using the collected raw data, we additionally prepared two more types of dataset, which are in ratio and binary format respectively. Dividing away-team's records by the records of the corresponding home-team generated the ratio dataset, while the binary dataset was obtained by comparing the record values. We applied seven classification techniques to three (raw, ratio, and binary) datasets. The employed data mining techniques are decision tree, random forest, logistic regression, neural network, support vector machine, linear discriminant analysis, and quadratic discriminant analysis. Among 21(= 3 datasets${\times}$7 techniques) prediction scenarios, the most accurate model was obtained from the random forest technique based on the binary dataset, which prediction accuracy was 84.14%. It was also observed that using the ratio and the binary dataset helped to build better prediction models than using the raw data. From the capability of variable selection in decision tree, random forest, and stepwise logistic regression, we found that annual salary, earned run, strikeout, pitcher's winning percentage, and four balls are important winning factors of a game. This research is distinct from existing studies in that we used three different types of data and various data mining techniques for win-loss prediction in Korean professional baseball games.

• Journal of Korea Multimedia Society
• /
• v.11 no.12
• /
• pp.1635-1648
• /
• 2008

The extraction of important features in cancer cell image analysis is a key process in grading renal cell carcinoma. In this study, we applied three-dimensional (3D) texture feature extraction methods to cell nuclei images and evaluated the validity of them for computerized cell nuclei grading. Individual images of 2,423 cell nuclei were extracted from 80 renal cell carcinomas (RCCs) using confocal laser scanning microscopy (CLSM). First, we applied the 3D texture mapping method to render the volume of entire tissue sections. Then, we determined the chromatin texture quantitatively by calculating 3D gray-level co-occurrence matrices (3D GLCM) and 3D run length matrices (3D GLRLM). Finally, to demonstrate the suitability of 3D texture features for grading, we performed a discriminant analysis. In addition, we conducted a principal component analysis to obtain optimized texture features. Automatic grading of cell nuclei using 3D texture features had an accuracy of 78.30%. Combining 3D textural and 3D morphological features improved the accuracy to 82.19%. As a comparative study, we also performed a stepwise feature selection. Using the 4 optimized features, we could obtain more improved accuracy of 84.32%. Three dimensional texture features have potential for use as fundamental elements in developing a new nuclear grading system with accurate diagnosis and predicting prognosis.

• Journal of Preventive Medicine and Public Health
• /
• v.17 no.1
• /
• pp.5-24
• /
• 1984

There has been an immense need for elaborate studies on the complications and the neuological sequelae generated by acute carbon monoxide (CO) poisoning which is highly prevalent in Korea due to widespread adoption of the anthracite coal briquette as domestic fuel for heating and for cooking. For this epidemiological study, a total of 444 subjects who received hospital emergency care for acute CO poisoning during the period of March 1982 to February 1983 were randomly selected from the emergency patients's lists of 13 general hospitals in Seoul area. Informations on the neurological sequelae were elucidated by means of home visiting with prearranged questionnaire consisting questions and concise neurological examination. The findings obtained were summarized as follows; 1. The complications were found in 18% of the surveyed and acute decubitus was comprised 67.5% of the complications. 2. The total cumulative incidence of the neurological sequelae was 41.2 per 100 patients and the absolute incidence rate regardless of the duration after poisoning was 40.8%. 3. The incidence of the neurological sequelae was higher in the older age than in the younger and also higher in female than in male. Twice higher incidence was observed in the admitted patients than in the non-admitted patients and the incidence became higher in proportion to the duration of CO exposure, coma and admission. The poorer the consciousness level of patients found, at emergency room and at discharge, the higher the incidence. The incidence of the neurological sequelae by emergency care was higher in hyperbaric oxygen therapy group(51.9%) than in 100% $O_2$ group(38.0%) 4. A total of five variables significantly associated with the occurrence of the neurological sequelae were selected by the stepwise discriminant analysis. The variables were following course of emergency care, age, consciousness level at discharge, admission duration, and consciousness level at emergency room in their sequence of discriminant power. Eight variables were selected as those associated with the degree of the neurological sequelae through the stepwise multiple regression analysis. Of these variables, the acute decubitus alone explained 21.1% of the total variation ana all the eight variables could explain 36.5% of the same. The remaining seven variables listed in the order of their relative importance were: age, consciousness level at discharge, admission duration, coma duration and consciousness level at emergency room. 5. It was postulated that unexpectedly high incidence of the neurological sequelae of the CO poisoning in this epidemiological study was mainly due to the inadequate emergency care and the lack of efficient and sophisticated treatment measure. In the effort to minimize the incidence of grave neurological sequelae of acute CO poisoning, new guidelines for the emergency care and treatment should be pursued with efficient ways.

• Korean Journal of Child Studies
• /
• v.17 no.1
• /
• pp.189-212
• /
• 1996

The purpose of this study was to investigate how parental childrearing behavior and children's sibling relationships are related to children's self-esteem. 440 middle class families consisting of two children and their parents were the subjects of this study. The research instruments included a childrearing behavior questionnaire, the Sibling Relationships Questionnaire (Furman & Buhrmester, 1985), and the Self-Perception Profile for Children (Harter, 1985). Analyses of the data included correlation analysis, canonical correlation, regression, stepwise multiple regression, and MANOVA with stepwise discriminant analysis as the follow-up test. The most powerful predictors of children's self-esteem were the Warmth-Acceptance of childrearing behavior and the Warmth-Closeness of sibling relationships. The self-esteem dimension was best predicted by parental childrearing behavior and by children's sibling relationships was Global Self-Worth. Behavioral Conduct was best predicted by the Rejection-Restriction factor of childrearing, and by Conflict (for boys) and Rivalry (for girls) factors of sibling relationships. Children's self-esteem was related more strongly to the Warmth-Acceptance and the Rejection-Restriction of opposite-sex parents. The effects of Permissiveness-Nonintervention were stronger in same-sex parent-child dyads. Parental childrearing behaviors accounted for boy's self-esteem better than girl's with the exception of Behavioral Conduct. Sibling relationships accounted for girl's self-esteem better than boy's. The $2{\times}2$ MANOVA revealed interaction effects of parental childrearing behaviors and sibling relationships on children's self-esteem. Two factors of Rivalry and Conflict in sibling relationships and all three factors of childrearing behaviors showed significant interaction effects, The childrearing factor of Permissiveness-Nonintervention and the sibling factor of Rivalry, which were relatively weak predictors of self-esteem when acting alone, gained power in explaining children's self-esteem within the interactional context.

• Korean Journal of Health Education and Promotion
• /
• v.26 no.3
• /
• pp.111-124
• /
• 2009

Objectives: The purpose of this study was to provide basic data for suggesting the realistic plan for the prevention of induced abortion. We performed an analysis of a recognition and actual conditions of the induced abortion according to the marital status. Methods: Participants were 681 persons of 20 years of age or older who lived in 7 cities and provinces. The data was collected with a structured questionnaire. For data analysis, Chi-square test, t-test and stepwise discriminant analysis were utilized. Results: The results were as follows. When degrees of recognition were analyzed according to the marital status, married were found to have higher levels of recognition of induced abortion than unmarried. From comparison of the recognition, there were significant differences between married and unmarried in terms of the recognition of the law, premarital sex, sex consciousness, intention of induced abortion, knowledge of contraception. Conclusion: The above results showed that marital status was strongly related to the recognition of induced abortion. And measures for preventing the induced abortion, specially about 20 percents of risky unmarried, should be focused.

• Korean Journal of Psychosomatic Medicine
• /
• v.11 no.2
• /
• pp.205-213
• /
• 2003

Objectives: The authors intended to evaluate long-term outcome of group cognitive behavioral therapy(GCBT) for panic disorder and examined the variables to predict high end-state functioning in 12 months follow-up. Methods: 236 patients meeting DSM-IV criteria for panic disorder were assessed by STAI, ASI, BDI, BSQ, PBQ, ACQ at pre & post treatment, and were asked about the frequency of panic attacks during recent one month. We executed stepwise discriminant analysis on the clinical variables at pre treatment assessment to find the variables for discriminating between high end-state function(HES) and low end-state function(LES). Results: After GCBT, 82.6% of panic patients maintained high end-state functioning at 12 months follow-up. The results of discriminant analysis showed that BDI and BSQ at pre treatment assessment were significant variables to predict end-state functioning at 12 months follow-up. The hit ratio of discriminant analysis was 69.2%. Conclusion: These results suggest that the therapeutic effect of GCBT can be maintained through 12 months. Especially, it is likely that patients who were less depressed and who had less experienced the fear of physical symptoms at pre-treatment can maintain HES.