• Journal of the Korean Data and Information Science Society
• /
• v.25 no.2
• /
• pp.305-315
• /
• 2014

In a simple linear regression, a linear relationship between an explanatory variable and a response variable can be easily recognized in the scatter plot of them. The lack of fit test for the replicated data is commonly used for testing the linearity but it is not easy to test the linearity when the explanatory variable is not replicated. In this paper, we propose three new test statistics for testing the linearity regardless of replication using the principle of average slope and validate them through several simulations and empirical studies.

• The Korean Journal of Applied Statistics
• /
• v.5 no.1
• /
• pp.29-39
• /
• 1992

A testing and estimation procedure is considered for changes at unknown time point in simple linear regression model. A test statistic of quadratic form is suggested. We also discuss the asymptotic distribution and its level control. The proposed method is compared with the likelihood ratio test through a example.

• Proceedings of the KIPE Conference
• /
• 2018.07a
• /
• pp.246-248
• /
• 2018

본 논문에서는 배터리 수명의 지표인 SOH(state of health) 추정 시 배터리 노화에 따라 방전 용량의 급격한 변화가 발생하면 SOH도 변화하게 된다. 이로 인해 잘못된 SOH의 정보를 가지고 오게 되며 배터리의 안정성 및 신뢰성에 문제가 된다. 본 논문에서는 방전 용량과 내부 저항의 선형적 관계를 확인하고, 방전 용량과 내부저항을 고려한 단순선형회귀모형(simple linear regression model)을 모델링하였다. 방전 용량의 급격한 변화나 오프라인 기반 방전 용량을 측정함에 어려움이 있는 경우 단순선형회귀모형에 따라 방전 용량을 추정하여 SOH를 보정하는 기법을 제안하고 이에 대한 검증을 수행하였다.

• Journal of the Korean Data and Information Science Society
• /
• v.20 no.2
• /
• pp.251-260
• /
• 2009

In this study, we propose a new estimation method based on autocovariance for selecting optimal estimators of the regression coefficients in the simple linear regression model. Although this method does not seem to be intuitively attractive, these estimators are unbiased for the corresponding regression coefficients. When the exploratory variable takes the equally spaced values between 0 and 1, under mild conditions which are satisfied when errors follow an autoregressive moving average model, we show that these estimators have asymptotically the same distributions as the least squares estimators. Additionally, under the same conditions as before, we provide a self-contained proof that these estimators converge in probability to the corresponding regression coefficients.

• The Journal of Engineering Geology
• /
• v.24 no.1
• /
• pp.101-110
• /
• 2014

This study presents the results of a regression analysis of the point-load strength ($I_{s(50)}$) and the uniaxial compressive strength (UCS) of granites in Korea. The regression was carried out for three cases using the least-squares method, reclassifying the granite samples based on their physical properties. The first regression analysis through the origin according to the weathering grade did not give a result with a sufficient degree of confidence, due to the small number of samples. However, the general trend of the correlation between UCS and $I_{s(50)}$according to weathering grade shows that the slope of the linear regression for weathered granite is steeper than that for fresh granite. The second analysis was a simple linear regression for all the granite samples using the least-squares method as well as a linear regression using the bootstrap resampling method in order to increase the confidence level and the accuracy of the regression results. The third regression considered the average strength of granite groups reclassified according to physical properties. These linear regression analyses yielded linear regression equations with slopes of 14 and small standard deviations being similar to values reported in previous studies on Korean granites, but whose intercept values range from 16 to 43 and have a larger standard deviation than those of the present study. In conclusion, it would be advisable to estimate UCS from $I_{s(50)}$, considering the error range derived from the deviation of the regression equations.

• Proceedings of the Korean Statistical Society Conference
• /
• 2004.11a
• /
• pp.265-270
• /
• 2004

불균형중첩오차구조를 갖는 단순선형회귀모형에서 나타나는 두 분산의 합에 대한 신뢰구간을 구하기 위하여 Ting et al.(1990) 방법과 Graybill and Wang(1980) 방법과 Tsui and Weerahandi(1989)가 제안한 일반화 축량(generalized pivotal quantity)방법을 이용한 두 가지 방법 등 모두 네 가지 신뢰구간을 제안한다. 신뢰구간의 적절성을 판단하기 위하여 여러 가지 불균형 설계에 대하여 SAS/IML로 시뮬레이션을 실행하고 신뢰계수와 신뢰구간의 평균 길이를 비교한다. 불균형중첩오차구조를 갖는 단순선형회귀모형의 두 분산의 합에 대한 네 가지 신뢰구간들이 주샘플링 단위의 변화에 따라 어느 방법이 적절한 신뢰구간을 구축하는지 추천하고, 실제 예제를 적용하여 시뮬레이션의 결과와 일관성이 있는지를 확인한다.

• The Korean Journal of Applied Statistics
• /
• v.26 no.6
• /
• pp.953-958
• /
• 2013

This study suggests a measurement error model based on skewed normal distribution instead of normal distribution to identify slope parameter properties in a simple liner regression model. We prove that the slope parameter in a simple linear regression model is underestimated.

• Journal of Advanced Navigation Technology
• /
• v.24 no.2
• /
• pp.128-133
• /
• 2020

In this study, an 1-dimensional positioning method using iterative linear regression for path loss expression is proposed. In the proposed method, received signal strengths (RSS) measured in several locations and distances between the measuring locat ions obtained by dead reckoning are used to derive a linear regression for the path loss from the transmitting beacon. In the proposed method, for the distance between the transmitting beacon and a target measuring location, several tentative values are assumed. For each tentative value, a linear regression is obtained. Among the linear regression expressions, the one closest to the known reference RSS value is selected and used to derive the distance to the target location. Test results show that the proposed method is more accurate than path loss model.

• Korean Journal of Soil Science and Fertilizer
• /
• v.31 no.4
• /
• pp.359-367
• /
• 1998

In order to determine the nitrogen supplying capabilities (NSC) of corn fields, 47 field experiments were performed in Pennsylvania over 3 year from 1986 and NSCs were estimated by the regression analysis with chemical properties and soil attributes. Although the content of $NO_3-N$ in soil showed the best correlation with NSC ($R^2=0.518$), the standardized partial regression coefficient of $NO_3-N$ for NSC was 0.52, with some variations over the years. This value was slightly higher than those of the other properties which ranged from 0.001 to 0.351. Multiple linear regression with soil attributes for the evaluation of NSC was better than simple regression with $NO_3-N$. The coefficient of determination ($R^2$) for the evaluation of NSC was gradually increased; 0.599 with selected chemical properties, 0.698 with quantitative attributes(chemical properties and depth of Ap horizon), and 0.839 with quantitative and selected qualitative soil attributes. Consequently, in order to evaluate NSC, analysis by multiple linear regression with soil attributes was more reliable and better model than by the simple regression model.

• Proceedings of the Korean Society of Disaster Information Conference
• /
• 2015.11a
• /
• pp.259-260
• /
• 2015

재해가 발생하기 전에 피해규모와 이에 따른 영향 및 피해액을 신속하게 추정하는 것은 효율적인 재난 관리를 하는데 있어 중요하고, 더불어 정책결정자들이 의사결정을 할 때 도움이 될 수 있다. 하지만 기존의 연구는 단순 재해 피해발생 후에 그 피해액 혹은 복구액을 산정하는 수준이며, 현재 피해액 추정에 대한 연구는 전무 하며 그 기술개발에 또한 초기단계에 있다. 이에 피해액 추정에 대한 연구가 시급한 실정이다. 실질적으로 자연재해 정보에 대한 수요가 급증하고 있는 반면에 체계적이고 일관된 통계기반의 정보체계는 미흡하다. 이에 국가에서도 재해에 대한 피해액 및 복구비 산정 등 예산편성에 있어 큰 어려움을 겪고 있다. 기존통계 방식에 대한 개선의 필요성 인식과 더불어 본 연구의 목적은 재해 발생 전에 그 피해규모와 영향을 고려하여 이에 따른 피해액을 신속하게 추정하기 위함에 있다. 본 연구에서는 비선형 회귀식을 이용하여 강우-홍수피해액에 대한 함수를 제시하고자 한다.