• Communications of the Korean Mathematical Society
• /
• v.20 no.1
• /
• pp.145-159
• /
• 2005

In this paper we consider the problem of testing statistical hypotheses for unknown parameters in nonlinear regression models and propose three asymptotically equivalent tests based on regression quantiles estimators, which are Wald test, Lagrange Multiplier test and Likelihood Ratio test. We also derive the asymptotic distributions of the three test statistics both under the null hypotheses and under a sequence of local alternatives and verify that the asymptotic relative efficiency of the proposed test statistics with classical test based on least squares depends on the error distributions of the regression models. We give some examples to illustrate that the test based on the regression quantiles estimators performs better than the test based on the least squares estimators of the least absolute deviation estimators when the disturbance has asymmetric and heavy-tailed distribution.

• Journal of Information Processing Systems
• /
• v.6 no.2
• /
• pp.235-252
• /
• 2010

Regression testing is a very costly process performed primarily as a software maintenance activity. It is the process of retesting the modified parts of the software and ensuring that no new errors have been introduced into previously tested source code due to these modifications. A regression test selection technique selects an appropriate number of test cases from a test suite that might expose a fault in the modified program. In this paper, we propose both a regression test selection and prioritization technique. We implemented our regression test selection technique and demonstrated in two case studies that our technique is effective regarding selecting and prioritizing test cases. The results show that our technique may significantly reduce the number of test cases and thus the cost and resources for performing regression testing on modified software.

• Journal of the Korean Data and Information Science Society
• /
• v.14 no.2
• /
• pp.153-165
• /
• 2003

This paper considers the likelihood ratio test statistic based on nonlinear regression quantiles estimators in order to test of hypothesis about the regression parameter $\theta_o$ and derives asymptotic distribution of proposed test statistic under the null hypothesis and a sequence of local alternative hypothesis. The paper also investigates asymptotic relative efficiency of the proposed test to the test based on the least squares estimators or the least absolute deviation estimators and gives some examples to illustrate the application of the main result.

• Journal of applied mathematics & informatics
• /
• v.8 no.2
• /
• pp.669-682
• /
• 2001

In this paper a nonparametric test for the parallelism of k regression lines against ordered alternatives, when the independent variables are positive and all regression lines have a common intercept is proposed. The proposed test is based on a Jonckheere-type statistic applied to residuals. Under some conditions the proposed test statistic is asymptotically distribution-free. The small-sample powers of our test are compared with other tests by a Monte Carlo study. The simulation results show that the proposed test has significantly higher empirical powers than the other tests considered in this paper.

• Journal of Information Technology Services
• /
• v.14 no.2
• /
• pp.129-142
• /
• 2015

The purpose of this study is to suggest an effective regression testing method in order to minimize the scope of test resulting from the modification of software and to prevent mismatch of test case and test objects. As a way to improve the efficiency of regression testing which uses a change-centric testing technique, the method flow is analyzed and grasped through a static analysis based on source code in order to identify modified parts. After the order of priority is set according to the results of user action log-based dynamic analysis on identified regression testing objects, test effect can be raised by adjusting the order of priority using code complexity. Quality assurance coverage can be checked using the user action log suggested in this study, and the progress of test and whether or not each function has been verified can be checked, too. In addition, by minimizing test parts and adjusting the order of test, costs and time can be saved, making it possible to conduct regression testing effectively.

• The Pure and Applied Mathematics
• /
• v.20 no.2
• /
• pp.109-116
• /
• 2013

In this paper, we propose a Jonckheere type test statistic for testing the parallelism of k regression lines against ordered alternatives. The order restriction problems could arise in various settings such as location, scale, and regression problems. But most of theory about the statistical inferences under order restrictions has been developed to deal with location parameters. The proposed test is an application of Jonckheere's procedure to regression problem. Asymptotic normality and asymptotic distribution-free properties of the test statistic are obtained under some regularity conditions.

• The Journal of Information Systems
• /
• v.4
• /
• pp.27-39
• /
• 1995

This study tests the fuzzy linear regression model to see if there is a performance difference between it and the classical linear regression model. These results show that FLR was better as f forecasting technique when compared with CLR. Another important find in the test of the two different regression methods is that they generate two different predicted P/E ratios from expected value test, variance test and error test of two different regressions, though we can not see a significant difference between two regression models doing test in error measurements (GMRAE, MAPE, MSE, MAD). So, in this financial setting we can conclude that FLR is not superior to CLR, comparing and testing between the t재 different regression models. However, FLR is better than CLR in the error measurements.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.8
• /
• pp.157-164
• /
• 2021

Regression analysis is a well-known statistical technique useful to explain the relationship between response variable and predictor variables. In particular, Researchers are interested in comparing the regression coefficients(intercepts and slopes) of the models in two independent populations. The Chow test, proposed by Gregory Chow, is one of the most commonly used methods for comparing regression models and for testing the presence of a structural break in linear models. In this study, we propose the use of permutation method and compare it with Chow test analysis for testing the equality of two independent linear regression models. Then simulation study is conducted to examine the powers of permutation test and Chow test.

• Communications for Statistical Applications and Methods
• /
• v.19 no.3
• /
• pp.471-478
• /
• 2012

In this paper, we present a permutation test to select the number of pairs of canonical variates in canonical correlation analysis. The existing chi-squared test is known to be limited to normality in use. We compare the existing test with the proposed permutation test and study their asymptotic behaviors through numerical studies. In addition, we connect canonical correlation analysis to regression and we we show that certain inferences in regression can be done through canonical correlation analysis. A regression analysis of real data through canonical correlation analysis is illustrated.

• Communications for Statistical Applications and Methods
• /
• v.15 no.5
• /
• pp.709-717
• /
• 2008

The difference of two one-sided kernel estimators is usually used to detect the location of the discontinuity points of regression function. The large absolute value of the statistic imply discontinuity of regression function, so we may use the difference of two one-sided kernel estimators as the test statistic for testing null hypothesis of a smooth regression function. The problem is, however, we only know the asymptotic distribution of the test statistic under $H_0$ and we hardly expect the good performance of test if we rely solely on the asymptotic distribution for determining the critical points. In this paper, we show that if we adjust the bias of test statistic properly, the asymptotic rules hold for even small sample size situation.