• The Korean Journal of Applied Statistics
• /
• v.19 no.3
• /
• pp.407-419
• /
• 2006

In this paper, we analyze household overdue loans in Korea which has been causing serious social and economical problems. We consider customers of Bank A in Korea and focus on overdue cash services which have been snowballing in the past few years. From analysis of overdue loans, one can predict possible delays for current customers as well as build a credit evaluation and risk management system for future customers. As a statistical analytical tool, we propose a two-stage Generalized Linear regression Model (GLM) which assumes a logistic model for presence/non-presence of overdue and a gamma model for the amount of overdue in the case of overdue. We perform goodness of fit test for the two-stage model and select significant explanatory variables in each stage of the model. It turns out that age, the amount of credit loans from other financial companies, the amount of cash service from other companies, debit balance, the average amount of cash service, and net profit are important explanatory variables relevant to overdue credit card cash service in Korea.

• Journal of Korean Society of Transportation
• /
• v.17 no.5
• /
• pp.19-32
• /
• 1999

In this study, the trip characteristics of workers in the city are systematically analyzed. The trip behaviors and socioeconomic characteristics of workers are analyzed using Person Trip Survey Data of 1988 and 1992 in Taegu Metropolitan area. With the results of behavioral analyses, the daily travel pattern of workers is shown as one tour contained two trips and it is relatively simple and stable. Also the rate using the same mode in a day is Presented as high ratio. So, it can be explained that the choice of worker\`s first trip is fixed his/her travel mode for his/her daily travel mode. Based on these analyses, the mode choice model for workers is developed by applying the Multi-nominal Logit Model with the choice set of bus, taxi, and car. The explanatory variables of this model include sex, age, auto, travel time, and cost. Empirical tests of the model show encouraging results. After that, the temporal transferability of the model is examined by the Pairwise t-test and five indexes far the model of 1988 and 1992. The results of examination are satisfied with each significance level of the explanatory variables and five indexes. Therefore. it can be concluded that the temporal transferability of this model developed in this study is resonable.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.12 no.1
• /
• pp.147-157
• /
• 2013

This study developed the transit transfer mode choice model aimed Daegu transit users using multinomial logit model. Dependent variables of estimating multinomial logit model were transit transfer modes such as bus to bus, bus to subway, subway to subway, bus to others, and subway to others, and explanatory variables which affect transit transfer mode choice were sex, age, occupation, handicap, transfer area, purpose of travel and travel time. Also probability regarding explanatory variables was estimated using multinomial logit model and limit marginal analysis was carried out according to explanatory variables(cost, time). In the results, indicating goodness of fit is very reasonable as ${\rho}^2$=0.354. According to the result of marginal analysis for the selection of probability, when travel time is increased, users of bus to bus and bus to subway prefer to use subway to subway. Furthermore users of bus to bus and bus to subway prefer to use bus to others and subway to others when travel cost is increased in the result of marginal analysis for the selection of probability.

• Journal of Digital Convergence
• /
• v.19 no.7
• /
• pp.271-282
• /
• 2021

This study was conducted to develop a model for predicting the length of stay for premature infants through machine learning. For the development of this model, 6,149 cases of premature infants discharged from the hospital from 2011 to 2016 of the discharge injury in-depth survey data collected by the Korea Centers for Disease Control and Prevention were used. The neural network model of the initial hospitalization was superior to other models with an explanatory power (R²) of 0.75. In the model added by converting the clinical diagnosis to CCS(Clinical class ification software), the explanatory power (R²) of the cubist model was 0.81, which was superior to the random forest, gradient boost, neural network, and penalty regression models. In this study, using national data, a model for predicting the length of stay for premature infants was presented through machine learning and its applicability was confirmed. However, due to the lack of clinical information and parental information, additional research is needed to improve future performance.

• Journal of Korea Technology Innovation Society
• /
• v.1 no.3
• /
• pp.374-385
• /
• 1998

This study is estimates telephone service demand based on empirical studies of telecommunication service demand model. First, the telephone charge(call price index) by each location and subscription fee bring about a negative effect to telephone distribution rate: while the other explanatory variables bring about a positive effect. Second, the flexibility of telephone charge in A location(relevant location) and the flexibility between the distance of A location and B location are negative values, while the flexibility of other explanatory variables is represented in a positive value. This means that the long distance call numbers from A location to B location are in inverse proportion against the phone charge(call price index) of A location and against the distance between A location and the distance of other locations except A location, while they are in direct proportion with an average call number per minute from A location to other locations except A location, and also with subscription numbers of A location, other subscribers in locations other than A location, and the total expenditures of A location.

• ETRI Journal
• /
• v.28 no.4
• /
• pp.541-544
• /
• 2006

From its foundation until 2004, ETRI has registered over 1,000 US patents. This letter analyzes the characteristics of these patents and addresses the explanatory factors affecting their citation counts. For explanatory variables, research team related variables, invention specific variables, and geographical domain related variables are suggested. Zero-altered count data models are used to test the impact of independent variables. A key finding is that technological cumulativeness, the scale of invention, outputs in the electronic field, and the degree of dependence on the US technology domain positively affect the citation counts of ETRI-invented US patents. The magnitude of international presence appears to negatively affect the citation counts of ETRI-invented US patents.

• Communications for Statistical Applications and Methods
• /
• v.3 no.1
• /
• pp.73-85
• /
• 1996

In this paper, we consider the distribution-free confidence intervals for the reliability of the stress-strength model when the stress X and strength Y depend linearly on some explanatory variables z and w, respectively. We apply these confidence intervals to the Rocket-Motor data and compare the results to those of Guttman et al. (1988). Some simulation results show that the distribution-free confidence intervals have better performance for nonnormal errors compared to those of Guttman et al. (1988), which are designed for normal random variables in respect that the former yield the coverage levels closer to the nominal coverage level than the latter.

• Journal of Family Resource Management and Policy Review
• /
• v.6 no.1
• /
• pp.103-115
• /
• 2002

The purpose of this study was to analyze which factor, influenced the business and family performance success. Data were obtained from 248 family households. Proxy variable of the business performance was gross business income and of the family performance was the Family AFGAR scores. The multiple regression analysis was conducted for both the business performance equation and family performance equation. The main results of this study were as followings: The results indicated the effects of various business and family characteristics on performance and their contributions to the business and family performance model. Nine explanatory variables such as sex, being home-based, number of hours worked per week, number of family employee, number of nonfamily employee, total asset, the presence of young child under 6, nonbusiness income, and role conflicts were statistically significant in the business performance equation and three explanatory variables such as the hours worked per week, family stress scores, and role conflicts were statistically significant in the family performance equation. The results indicated the need for a more comprehensive view of family business performance.

• The Korean Journal of Applied Statistics
• /
• v.11 no.2
• /
• pp.451-459
• /
• 1998

The analysis of data from a multiresponse experiment requires careful consideration of the multivariate nature of the data. In a multiresponse sitation, the optimization problem is more complex than in the single response case. The biplot is a graphical tool which make the analyst to understand the correlation of the response variables, the relation of the response variables arid the explanatory variables and the relative importance of the explanatory variables. In case of good fitting of the first order model, we can draw the biplot with the first order experimental design. Otherwise, we can make the biplot with the second order experimental design by adding other experimental points.

• Journal of Digital Convergence
• /
• v.10 no.9
• /
• pp.111-122
• /
• 2012

In this study, we consider the integration of cognitive components and emotion to account for communication behaviors and activism on a consumer advocacy website. The challenge of integrating situational theory and anger activism model was empirically explored utilizing an online survey with the members of the virtual community, which was designed to raise issues and to protest against the product defects of a product. Our findings indicated that along with the cognitive perception in the situational theory, anger as a negative emotion was the most significant predictor strongly associated with communication behaviors and activism on the organization. More interestingly, the model that integrates anger with cognitive components significantly improves its explanatory power compared with one including only cognitive components as explanatory variables.