• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.35 no.5
• /
• pp.415-422
• /
• 2017

Finding influential factors from given clustering result is a typical data science problem. Genetic Algorithm based method is proposed to derive influential factors and its performance is compared with two conventional methods, Classification and Regression Tree (CART) and Chi-Squared Automatic Interaction Detection (CHAID), by using Dunn's index measure. To extract the influential factors of preference towards political parties in South Korea, the vote result of $18^{th}$ presidential election and 'Demographic', 'Health and Welfare', 'Economic' and 'Business' related data were used. Based on the analysis, reverse engineering was implemented. Implementation of reverse engineering based approach for influential factor analysis can provide new set of influential variables which can present new insight towards the data mining field.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.15 no.8 s.101
• /
• pp.939-944
• /
• 2005

In this paper, a sound quality index for the evaluation of the vehicle intake noise was developed through a correlation analysis of the objective measurement and the subjective evaluation. First, intake orifice noise was measured at the wide-open throttle sweep condition. And then, the acoustic transfer function between intake orifice noise and interior noise was measured. Simultaneously, subjective evaluation was carried out with a 10-scale score by 8 special engineers. The correlation analysis between the psychoacoustic parameters derived from the measurement and the subjective evaluation was performed. The most critical factor was determined and the corresponding sound quality index for intake noise was obtained from the multiple factor regression analysis method. Finally, the effectiveness of the proposed index was validated.

• Journal of the Korean Home Economics Association
• /
• v.33 no.4
• /
• pp.49-64
• /
• 1995

The purpose of this study was to develop the Daily Hassles Scale for children in Korea. The subject were 444 children of 184 fourth graders and 260 sixth graders selected form five elementary schools in Seoul(217 male and 227 female). A questionnaire consisting of 90-item daily hassles scale, demographic questions, and some additional questions was used as a methodological instrument. statistics used for data analysis were X2, cramer's V, factor analysis, multi-regression, Pearson's r, Cronbach's α. The major findings of this study were as follows. 1) 87 items of the 90-item scale were acceptible through item discriminant method. The discriminant coefficients of the items(Cramer's V) ranged form .28 to .73. 2) 6 factors(parents, home environment, friends, studies, teachers & school, the surroundings) were extracted from factor analysis. Multi-regression analysis conducted to reduce the length of scale have drawed 42 items for 'the Daily Hassles Scale for Children in Korea'. The correlation between this scale and the Quality of Life Scale(Olson & Barnes, 1982) was conducted to test the criterion-related validity, and the coefficient was significant(r=-.52, p<.001).3) Finally, reliability coefficients(Cronbach'α) of this scale was. 85.

• Journal of Families and Better Life
• /
• v.16 no.4
• /
• pp.155-166
• /
• 1998

The purpose of this study was to develop ' the Behavior Problem Scale for Children and Adolescence' The 518 subjects were selected from 5th and 6th grades of elementary schools and first and second grades of middle schools in Pusan. Statistics used for data analysis were χ2 cramer's V, factor analysis multi-regression Pearson's r, Cronbach's a. The major finding of this study were as follows 1) 80 items of the 159 item scale were acceptable through item discriminant method The discriminant coefficients of the items(Cramer's V) ranged from .48 to .81. 2) 6 factors(shyness aggression hyperactivity withdrawal anxious immature) extracted from factor analysis,. Multi-regression analysis conducted to reduce the length of scale have drawn 42 items for 'the Behavior Problem Scale Children and Adolescence' 3) Reliability coefficients(Cronbach's a) of this scale was 94.

• 한국방재학회:학술대회논문집
• /
• 2008.02a
• /
• pp.711-714
• /
• 2008

The objective of this study is to analyze the relationship between the watershed characteristics and the critical duration of design rainfall. For estimation of critical duration, adjustment Huff's method and ILLUDAS urban runoff model were applied to urban 21 areas. Watershed characteristics such as area, channel length, channel slope, shape factor, and pipe density were used to simulate correlation analysis. The conclusions of this study are as follows; it is revealed that critical duration is influenced by the watershed characteristics such as pipe density, area and channel length. Also, multiple regression analysis using watershed characteristics is carried out and the determination coefficient of multiple regression equation shows 0.972.

• Journal of Korean Society for Quality Management
• /
• v.24 no.2
• /
• pp.128-141
• /
• 1996

The advent of the quality-improvement movement caused a great expansion in the use of statistically designed experiments in industry. The regression method is often used for the analysis of data from such experiments. However, the data for a quality characterstic often takes the form of counts or the ratio of counts, e.g. fraction of defectives. For such data the analysis using generalized linear models is preferred to that using the simple regression model. In this paper we introduce the generalized linear model and show how it can be used for the analysis of non-normal data from quality-improvement experiments.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.10 no.3
• /
• pp.233-243
• /
• 2008

A tunnel behavior predicted in the investigation and design stage is often different from its actual behavior due to mainly the complexity of ground conditions. In a tunnel construction, therefore, it is necessary to ensure the stability of the tunnel by predicting the behaviors of the ground and the supports through observations and measurements, and modifying immediately excavation and reinforcing methods when necessary. To do so, it is important to be able to predict the final tunnel behavior based on the initial tunnel behavior as early as possible. In this study, the correlations were obtained between the initial and the final convergence by analyzing statistically the convergence measurement data, collected from two domestic road tunnels under construction using NATM. In order to estimate the unknown displacements, occurred during the period between the excavation and the first measurement, two methods were used - one is the method by means of regression analysis using a modified exponential function and the other the method by a simple linear regression analysis using the data measured within the distance from tunnel face equal to the tunnel diameter (D). Finally, the relationships were obtained between the initial and final convergence, including the non-measured displacements estimated from the two different methods, by performing linear regression analyses. The regression analysis results showed that there are clear linear relationships between the initial and final convegence and the difference between the two linear regression equations was not that large for when using the exponential function and the simple linear function to estimate the non-measured displacements.

• Journal of the East Asian Society of Dietary Life
• /
• v.23 no.6
• /
• pp.862-870
• /
• 2013

This study examines the effect of communication type of food service employees on the self-efficiency and organizational commitment. To accomplish the purpose of the study, a survey was conducted to employees of food service companies located in Seoul, from August 5, 2013 to August 25, 2013 as a method of empirical study. The results of this study was analyzed using the SPSS WIN 12.0 and statistical reliability analysis, factor analysis, frequency analysis, correlation analysis and multiple regression analysis were conducted for hypothesis verification. The analysis results of this study are as follows. First, the R-square of regression result for 'communication type will have a positive effect on self-efficiency' was .220. Also, it was found that factors of consideration (Beta=.286, t=5.622), instruction (Beta=.303, t=5.970) and business (Beta=.214, t=4.217) had positive effects on self-efficiency. Values for factors of consideration and instruction were higher than business. Second, the R-square of regression result for 'communication type will have a positive effect on organizational commitment' was .429. Also, it was found that factors of consideration (Beta=.453, t=10.428), instruction (Beta=.380, t=8.749) and business (Beta=.282, t=6.477) had positive effects on organizational commitment. Values for factors of instruction and business were high but value of consideration factor turned out the highest. Third, the R-square of regression result for 'self-efficiency will have a positive effect on organizational commitment' was .334. Also, it was found that out of the factors of positivity (Beta=.369, t=8.423), confidence (Beta=.415, t=8.833) and negativity (Beta=-.072, t=-1.536), positivity and confidence had positive effects on organization commitment, and only positivity and confidence showed high values in organization commitment. Upon analyzing the effect of factors of self-efficiency (positivity, confidence, negativity) on organizational commitment, it was found that positivity and confidence showed high relevance but the negativity factor did not have any correlation.

• Journal of Korea Water Resources Association
• /
• v.45 no.10
• /
• pp.1023-1034
• /
• 2012

Linear regression technique has been used widely in water resources field as well as various fields such as economics and statistics, and so on. Using fuzzy regression technique, it is possible to quantify uncertainty and reflect them to the regression model. In this study, fuzzy regression model is developed to compute design floods in any place in Pyeongchang River basin. In ungaged basins, it is usually difficult to obtain data required for flood discharge analysis. In this study, basin characteristics elements are analyzed spatially using GIS and the technique of estimating design flood in ungaged mountainous basin is studied based on the result. Fuzzy regression technique is applied to Pyeongchang River basin which has mountainous basin characteristics and well collected rainfall and runoff data through IHP test basin project. Fuzzy design flood estimation equations are developed using the basin characteristics elements for Pyeongchang River basin. The suitability of developed fuzzy equations are examined by comparing the results with design floods computed in 9 locations along the river. Using regional regression method and fuzzy regression analysis, the uncertainties of the design floods occurred from the data monitoring can be quantified.

• Proceedings of the KSRS Conference
• /
• v.2
• /
• pp.574-577
• /
• 2006

In this paper, we developed a forest fire detection algorithm which uses a regression function between NDVI and land surface temperature. Previous detection algorithms use the land surface temperature as a main factor to discriminate fire pixels from non-fire pixels. These algorithms assume that the surface temperatures of non-fire pixels are intrinsically analogous and obey Gaussian normal distribution, regardless of land surface types and conditions. And the temperature thresholds for detecting fire pixels are derived from the statistical distribution of non-fire pixels’ temperature using heuristic methods. This assumption makes the temperature distribution of non-fire pixels very diverse and sometimes slightly overlapped with that of fire pixel. So, sometimes there occur omission errors in the cases of small fires. To ease such problem somewhat, we separated non-fire pixels into each land cover type by clustering algorithm and calculated the residuals between the temperature of a pixel under examination whether fire pixel or not and estimated temperature of the pixel using the linear regression between surface temperature and NDVI. As a result, this algorithm could modify the temperature threshold considering land types and conditions and showed improved detection accuracy.