• Journal of the Korean Data and Information Science Society
• /
• v.20 no.4
• /
• pp.673-684
• /
• 2009

Lower Prices are offered through sales by telemarketing. This is to serve our customers by the fastest and most appropriate referral product that is most important to attract insurance. Therefore, Considering the time the customer's preferred products and preferred customer for screening and targeting, depending on what is the difference between the premiums. This study of the logistic regression model using datamining techniques, the life insurance companies in outbound telemarketing to support sales of the effect you want to validate. To join existing life insurance companies for the customer response and sales strategy based on the L segment and by age group, family-love insurance, accident insurance, and cancer insurance were in progress for the modeling. Set model based on the progress of the campaign to existing customers marketing methods and how to extract and run the model results has proven the superiority of the model. In addition, over time, depending on the aging model is set to a decline in operating profit to maximize the profits th update the model which was derived.

• Journal of Environmental Science International
• /
• v.32 no.3
• /
• pp.161-171
• /
• 2023

The present study evaluated growth factors affecting oak community plantations through literature review and a field survey. Specifically, 41 related literature sources were analyzed and field surveys were conducted to collect growth data. Previous studies were analyzed to identify variables with high frequency of use. The frequency of use was in the order of tree size > environment > planting density > forest age. Analysis of factors impacting height and diameter growth revealed that the growth rate of species other than Quercus variabilis was negative in the field survey. This may be because of differences between the actual trees planted and specifications in the construction drawings, which may be attributed to the site conditions and decisions made by the project subject during construction. Furthermore, simple linear regression analysis was conducted with time, height at planting, density, and species code as the independent variables and growth rate as the dependent variable. A strong positive linear correlation was noted between height and diameter. This work builds a foundation for developing a forest restoration model and simulation program based on a regression model derived from the four variables tested.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.21 no.5
• /
• pp.400-407
• /
• 2011

The suggested method of previous Son's study dichotomized subjective response data to modeling noise exposure-response. The method used maximum liklihood estimation instead of least square estimation and the noise exposure-response curve of the study was logistic regression analysis result. The method was originated to modeling community response rate such as %HA or %A. It can be useful when the subjective response was investigated based on predicted noise level. It is difficult to measure the single source emitting noise such as railway because various traffic noise sources combined in our life. The suggested method was adopted to model in this study and railway noise-exposure response curves were modeled because the noise level of this area was predicted data. The data of this study was used by previous Ko's paper but he dealt the area as combined noise area and divided the data by dominant noise source. But this study used all data of this area because the annoyance response to railway noise was higher than other noise according to the result of correlation analysis. The trend of the %HA and %A prediction model to train noise of this study is almost same as the model based on measured noise of previous Lim's study although the investigated areas and methods were different.

• Journal of the Korea Safety Management & Science
• /
• v.16 no.3
• /
• pp.257-266
• /
• 2014

This study was planned to investigate the effect of the exposure to hazard factors on work environment satisfaction. Existing researches about job satisfaction have focused on the general working conditions, such as working hours, wage, human relationship, job task and so on. Korean Working Conditions Survey was used for this study because that relevant questions were included. The effect of the exposure to hazard factors on work environment satisfaction may be produced by hierarchical regression analysis because of comparison with existing model for work environment satisfaction. The exposure to hazards factors were statistically significant effect on work environment satisfaction after adjusting other confounding variables, such as gender, age, educational level, job security, work hour, work load, work autonomy, social support, etc. This study has some limitation because that KWCS was cross sectional survey. Some researches about the causal effect and its mechanism may be suggested as future study.

• Journal of Fashion Business
• /
• v.19 no.5
• /
• pp.77-92
• /
• 2015

The purpose of this study is to examine and compare effects of psychological factors on market mavenism and fashion leadership in order to determine the differences of two influential groups in the marketplace. The data were collected from 20's-50's consumers through an online survey institute and a total of 857 questionnaires were analyzed. Demographic variables (gender, age, and income level) were entered into the regression model 1 as independent variables, and 6 factors of consumer self-confidence, clothing involvement, status consumption, and price consciousness were entered into the regression model 2. In the regression model 1, gender (female) alone was significant in explaining market mavenism, while the income level had a positive relationship with fashion leadership. In the regression model 2, information acquisition, social outcome, persuasion knowledge among consumer self-confidence, and status consumption were significant predictors of market mavenism. On the other hand, personal outcome, social outcome, persuasion knowledge, clothing involvement, and status consumption had an effect on the fashion leadership. When comparing magnitudes of effects in predicting market mavenism and fashion leadership, social outcome and status consumption showed to have stronger impacts on fashion leadership than on market mavenism. Psychological factors showed to be more powerful in predicting market mavenism or fashion leadership, as compared to demographic variables.

• Journal of Digital Convergence
• /
• v.13 no.6
• /
• pp.125-132
• /
• 2015

It is current situation that a number of issues are being raised how the weight is calculated from customer satisfaction survey. This study investigated how the weight of satisfaction for each quantile is different by comparing ordinary least square regression model to quantile regression model and carried out bootstrap verification to find the influence difference of regression coefficient for each quantile. As the analysis result of using R(Quantreg package) that is open software, it appeared that there was the influence size of satisfaction factor along study result and quantile and there was the significant difference statistically regarding regression coefficient for each quantile. So, to use quantile regression model that offers the influence of satisfaction factor for each customer group along satisfaction level would contribute to plan the quantitative convergence policy for customer satisfaction.

• Proceedings of the Korean Geotechical Society Conference
• /
• 2005.03a
• /
• pp.185-190
• /
• 2005

In this study, a probabilistic prediction model for debris flow occurrence was developed using a logistic regression analysis. The model can be applicable to metamorphic rocks and granite area. In order to develop the prediction model, detailed field survey and laboratory soil tests were conducted both in the northern and the southern Gyeonggi province and in Sangju, Gyeongbuk province, Korea. The six landslide triggering factors were selected by a logistic regression analysis as well as several basic statistical analyses. The six factors consist of two topographic factors and four geological and geotechnical factors. The model assigns a weight value to each selected factor. The verification results reveal that the model has 86.5% of prediction accuracy. Therefore, it is possible to predict landslide occurrence in a probabilistic and quantitative manner.

• Journal of Korea Planning Association
• /
• v.54 no.3
• /
• pp.106-118
• /
• 2019

In order to develop pedestrian navigation service that provides optimal pedestrian routes based on pedestrian satisfaction levels, it is required to develop a prediction model that can estimate a pedestrian's satisfaction level given a certain condition. Thus, the aim of the present study is to develop a pedestrian satisfaction prediction model based on three machine learning algorithms: Logistic Regression, Random Forest, and Artificial Neural Network models. The 2009, 2012, 2013, 2014, and 2015 Pedestrian Satisfaction Survey Data in Seoul, Korea are used to train and test the machine learning models. As a result, the Random Forest model shows the best prediction performance among the three (Accuracy: 0.798, Recall: 0.906, Precision: 0.842, F1 Score: 0.873, AUC: 0.795). The performance of Artificial Neural Network is the second (Accuracy: 0.773, Recall: 0.917, Precision: 0.811, F1 Score: 0.868, AUC: 0.738) and Logistic Regression model's performance follows the second (Accuracy: 0.764, Recall: 1.000, Precision: 0.764, F1 Score: 0.868, AUC: 0.575). The precision score of the Random Forest model implies that approximately 84.2% of pedestrians may be satisfied if they walk the areas, suggested by the Random Forest model.

• Journal of the Korea Computer Industry Society
• /
• v.10 no.5
• /
• pp.213-220
• /
• 2009

Regression analysis to estimate the fitted models and test hypotheses are basic statistical tools for survey data as well as experimental data. Data is collected as pairs of independent and dependent variables, and statistics are computed using matrix calculation. To estimate a best fitted model is a key to maximize reliability of regression analysis. To fit a regression model, plot data on XY axis and select the most fitted models. Researchers estimate the best model and test hypothesis with MS Excel's graph menu and matrix computation functions. In this study, I develop macros to estimate the fitted regression model and test hypotheses of relationship between variables. Subway electric charges data with one dependent variable and three independent variables are tested using developed macros, and compared with the results using built-in Excel of regression analysis.

• Journal of Biosystems Engineering
• /
• v.39 no.3
• /
• pp.151-157
• /
• 2014

Purpose: In order to develop strategies to prevent farm-work accidents relating to agricultural machinery, influential factors were examined in this paper. The effects of these factors were quantified using logistic regression. Methods: Based on the results of a survey on farm-work accidents conducted by the National Academy of Agricultural Science, 21 tentative independent variables were selected. To apply these variables to regression, the presence of multicollinearity was examined by comparing correlation coefficients, checking the statistical significance of the coefficients in a simple linear regression model, and calculating the variance inflation factor. A logistic regression model and determination method of its goodness of fit was defined. Results: Among 21 independent variables, 13 variables were not collinear each other. The results of a logistic regression analysis using these variables showed that the model was significant and acceptable, with deviance of 714.053. Parameter estimation results showed that four variables (age, power tiller ownership, cognizance of the government's safety policy, and consciousness of safety) were significant. The logistic regression model predicted that the former two increased accident odds by 1.027 and 8.506 times, respectively, while the latter two decreased the odds by 0.243 and 0.545 times, respectively. Conclusions: Prevention strategies against factors causing an accident, such as the age of farmers and the use of a power tiller, are necessary. In addition, more efficient trainings to elevate the farmer's consciousness about safety must be provided.