• Journal of Korean Society for Atmospheric Environment
• /
• v.26 no.5
• /
• pp.533-542
• /
• 2010

The $PM_{10}$ concentrations in the underground should be monitored for the health of commuters on the underground subway system. Seoul Metro and Seoul Metropolitan Rapid Transit Corporation are measuring several air pollutants regularly. As for the measurement of $PM_{10}$ concentrations, instruments based on $\beta$-ray absorption method and gravimetric methods are being used. But the instruments using gravimetric method give us 20-hour-average data and the $\beta$-ray instruments can measure the $PM_{10}$ concentration every one hour. In order to keep the $PM_{10}$ concentrations under a healthy condition, the air quality of the underground platform and tunnels should be monitored and controlled continuously. The $PM_{10}$ instruments using light scattering method can measure the $PM_{10}$ concentrations every less than one minute. However, the reliability of the instruments using light scattering method is still not proved. The purpose of this work is to study the reliability of the instruments using light scattering method to measure the $PM_{10}$ concentrations continuously in the underground platforms. One instrument using $\beta$-ray absorption method and two different instruments using light scattering method (LSM1, LSM2) were placed at the platform of the Jegi station of Seoul metro line Number 1 for 10 days. The correlation between the $\beta$-ray instrument and the LSM2 ($r^2$=0.732) was higher than that between the $\beta$-ray instrument and the LSM1 ($r^2$=0.393). Thus the LSM2 was chosen for further analysis. Three different regression analysis methods were tested: Linear regression analysis, Nonlinear regression analysis and Orthogonal regression analysis. When the instruments using light scattering method were used, the data measured these instruments have to be converted to actual $PM_{10}$ concentrations using some factors. With these analyses, the factors could be calculated successfully as linear and nonlinear forms with respect to the data. And the orthogonal regression analysis was performed better than the ordinary least squares method by 28.45% reduction of RMSE. These findings propose that the instruments using light scattering method light scattering method can be used to measure and control the $PM_{10}$ concentrations of the underground subway stations.

• Proceedings of the Korean Statistical Society Conference
• /
• 2000.11a
• /
• pp.187-191
• /
• 2000

In the analysis of repeated measurements, multivariate regression models that account for the correlations among the observations from the same subject are widely used. Like the usual univariate regression models, these multivariate regression models also need some model diagnostic procedures. In this paper, we propose a simple graphical method to detect outliers and to investigate the goodness of model fit in repeated measures data. The graphical method is based on the quantile-quantile(Q-Q) plots of the $X^2$ distribution and the standard normal distribution. We also propose diagnostic measures to detect influential observations. The proposed method is illustrated using two examples.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.13 no.9
• /
• pp.1935-1940
• /
• 2009

One of the key technical challenges of wireless sensor network (WSN) is location management of sensor nodes. Typical node location management methods use GPS, ultrasonic sensors or RSSI. In this paper we propose a new location management method which adopts regression analysis of RSSI measurement to improve the accuracy of sensor node position estimation. We also evaluated the performance of proposed method by comparing the experimental results with existing scheme. According to the results, our proposed method, LM-RAR, shows better accuracy than existing location management scheme using RSSI and Friis' equation.

• Proceedings of the KSRS Conference
• /
• v.2
• /
• pp.547-550
• /
• 2006

This study is to suggest a methodology to produce landslide hazard map by combining LRA (Logistic Regression Analysis) and AHP (Analytic Hierarchy Program) Approach. Topographic factors (slope, aspect, elevation), soil drain, soil depth and land use were adopted to classify landslide hazard areas. The method was applied to a 520 $km^2$ region located in the middle of South Korea which have occurred 39 landslides during 1999 and 2003. The suggested method showed 58.9 % matching rate for the real landslide sites comparing with the classified areas of high-risk landslide while LRA and AHP showed 46.1 % and 48.7 % matching rates respectively.

• Journal of Biomedical Engineering Research
• /
• v.22 no.4
• /
• pp.369-375
• /
• 2001

The statistical regression model is one of the most frequently used clinical analysis methods. It has basic assumption of linearity, additivity and normal distribution of data. However, most of biological data in medical field are nonlinear and unevenly distributed. To overcome the discrepancy between the basic assumption of statistical model and actual biological data, we propose a new analytical method based on artificial neural network. The newly developed multilayer perceptron(MLP) is trained with 120 data set (60 normal, 60 patient). On applying test data, it shows the discrimination power of 0.76. The diabetic risk factors were also identified from the MLP neural network model and the logistic regression model. The signigicant risk factors identified by MLP model were post prandial glucose level(PP2), sex(male), fasting blood sugar(FBS) level, age, SBP, AC and WHR. Those from the regression model are sex(male), PP2, age and FBS. The combined risk factors can be identified using the MLP model. Those are total cholesterol and body weight, which is consistent with the result of other clinical studies. From this experiment we have learned that MLP can be applied to the combined risk factor analysis of biological data which can not be provided by the conventional statistical method.

• Journal of Korea Water Resources Association
• /
• v.41 no.1
• /
• pp.17-25
• /
• 2008

Whether rating curves are used in practice or new ones are derived, the characteristics of regression analysis are often neglected. For example, a discharge rating curve, which is established from a regression of observed water levels (H) on observed flowrates(Q), is sometimes used for estimating a design water level corresponding to a simulated design flood runoff. However, if independent and dependent variables are changed with each other, the regression equation is changed in existing regression analysis, which is derived from vertical errors between observed data and regression line. Thus, regression equations should not be applied inversely. To avoid this problem, A new two-way variable least-squares regression analysis is proposed. The new method was applied to the rating curves of five water level stations on main stream of Nakdong River. The three kinds of regression models, which are respectively regression of Q versus H (model 1), H versus Q (model 2) and two-way (model 3), showed that the new method can reduce inadvertent mistakes when applied in practice.

• Journal of the Korean Data and Information Science Society
• /
• v.23 no.4
• /
• pp.643-648
• /
• 2012

There are lots of analysis to determine the relation between dependent variable and explanatory variables. Often the regression analysis is used to do this, and we can analyze the how much the explanatory variable can be related with dependent variable and how much the regression model can explain the data. But the validation check of regression model is usually determined by coefficient of determination. We should check the validation of regression coefficient with different methods. This paper introduces the method for validation check the regression coefficient using the jackknife regression and bootstrap regression in clinical data.

• Journal of the Korean Data and Information Science Society
• /
• v.2
• /
• pp.1-9
• /
• 1991

Shin, Tarumi and Tanaka(1989) discussed a method of sensitivity analysis in principal component regression(PCR) based on an influence function derived by Tanaka(1988). The present paper is its continuation. In this paper we first consider two new influence measures, then apply the proposed method to various data sets and discuss some properties of sensitivity analysis in PCR.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.23 no.4
• /
• pp.366-370
• /
• 2013

Regression Analysis is an analyzing method of regression model to explain the statistical relationship between explanatory variable and response variables. This paper introduce Theil's method to find a fuzzy regression model which explain the relationship between explanatory variable and response variables. Theil's method is a robust method which is not sensive to outliers. Theil's method use medians of rate of increment based on randomly chosen pairs of each components of ${\alpha}$-level sets of fuzzy data in order to estimate the coefficients of fuzzy regression model. We propose an example to show Theil's estimator is robust than the Least squares estimator.

• Communications for Statistical Applications and Methods
• /
• v.19 no.2
• /
• pp.293-301
• /
• 2012

Quantile regression proposed by Koenker and Bassett (1978) is a statistical technique that estimates conditional quantiles. The advantage of using quantile regression is the robustness in response to large outliers compared to ordinary least squares(OLS) regression. A regression tree approach has been applied to OLS problems to fit flexible models. Loh (2002) proposed the GUIDE algorithm that has a negligible selection bias and relatively low computational cost. Quantile regression can be regarded as an analogue of OLS, therefore it can also be applied to GUIDE regression tree method. Chaudhuri and Loh (2002) proposed a nonparametric quantile regression method that blends key features of piecewise polynomial quantile regression and tree-structured regression based on adaptive recursive partitioning. Lee and Lee (2006) investigated wage determinants in the Korean labor market using the Korean Labor and Income Panel Study(KLIPS). Following Lee and Lee, we fit three kinds of quantile regression tree models to KLIPS data with respect to the quantiles, 0.05, 0.2, 0.5, 0.8, and 0.95. Among the three models, multiple linear piecewise quantile regression model forms the shortest tree structure, while the piecewise constant quantile regression model has a deeper tree structure with more terminal nodes in general. Age, gender, marriage status, and education seem to be the determinants of the wage level throughout the quantiles; in addition, education experience appears as the important determinant of the wage level in the highly paid group.