• Journal of the Korean Data and Information Science Society
• /
• 제26권6호
• /
• pp.1513-1521
• /
• 2015

Selecting relevant variables for a statistical model is very important in regression analysis. Recently, variable selection methods using a penalized likelihood have been widely studied in various regression models. The main advantage of these methods is that they select important variables and estimate the regression coefficients of the covariates, simultaneously. In this paper, we propose a simple procedure based on a penalized h-likelihood (HL) for variable selection in Poisson hierarchical generalized linear models (HGLMs) for correlated count data. For this we consider three penalty functions (LASSO, SCAD and HL), and derive the corresponding variable-selection procedures. The proposed method is illustrated using a practical example.

• 응용통계연구
• /
• 제34권4호
• /
• pp.589-597
• /
• 2021

단순회귀와 다중회귀에서 회귀계수의 의미는 차이가 있고 회귀계수의 추정값은 같지 않을 뿐 아니라 그 부호가 서로 다른 경우도 발생한다. 회귀모형에서 설명변수의 상대적 기여도의 파악은 회귀분석의 수행의 중요한 부분이다. 표준화 회귀모형에서 표준화 회귀계수는 해당 설명변수를 제외한 나머지 설명변수의 값이 고정되어있는 상황에서 설명변수가 표준편차만큼 증가하였을 때 반응변수가 표준편차를 기준으로 얼마나 변화했는가로 해석할 수 있지만 표준화 회귀계수의 크기가 각 설명변수의 상대적 중요도를 나타내는 척도라고 할 수 없음은 잘 알려져 있다. 본 논문에서는 다중회귀에서 회귀계수의 추정량을 상관계수와 결정계수의 함수로 나타내고 이를 추가적인 설명력과 추가적인 결정계수의 관점에서 생각해 본다. 또한 다양한 산점도에서의 상관계수와 회귀계수 추정값의 관계를 알아보고 설명변수가 두 개인 경우에 구체적으로 적용해 본다.

• 한국지리정보학회지
• /
• 제20권3호
• /
• pp.181-194
• /
• 2017

본 연구의 목적은 산림재적 현장자료와 항공 LiDAR 자료 기반의 산림재적 추정을 위한 회귀모델의 개발이다. 추정 모델은 경상북도 봉화군 지역에서 임의추출법에 의해 선정된 30개의 원형 표본지로부터 산출한 표본지별 산림재적을 반응변수로 하고, 항공 LiDAR 원자료로부터 개별 표본지의 고도분포 백분위수(Height Percentiles, HP) 및 층위 단위 점 개체수 백분율(Height Bin, HB)을 추출하여 예측변수로 사용하여 구성하였다. 단순선형회귀분석, 이차 다항회귀분석 및 단계적 회귀분석 방법을 이용한 다중회귀분석을 실시하여 적합모델들의 후보들을 도출하였으며, 검증을 위하여 각 모델별로 교차 타당성 검증을 실시하여 PRESS 통계치를 구하였다. 모델의 $R^2$ 및 PRESS을 비교하여 적합성을 검토한 결과, $HB_{5-10}$, $HB_{15-20}$, $HB_{20-25}$, $HBgt_{25}$의 다중회귀모델의 $R^2$이 0.509로 가장 높고, $HP_{25}$ 단순회귀모델의 PRESS 값이 122.352으로 가장 낮은 것으로 나타났다. 수직구조가 복잡한 우리나라 산림재적을 추정하는 모델로는 다양한 수직적 정보를 포함하고 있는 $HB_{5-10}$, $HB_{15-20}$, $HB_{20-25}$, $HBgt_{25}$이 상대적으로 보다 적합하다고 사료된다.

• 한국물환경학회지
• /
• 제28권2호
• /
• pp.224-231
• /
• 2012

It is necessary to develop flow duration curve (FDC) on each unit watershed in order to analyze flow conditions in the stream for the management of Total Maximum Daily Loads (TMDLs). This study investigated a simple method to develop FDC for the general use of the curve. A simple equation for daily flow estimation was derived from the regression analysis between the 8-day interval flow data of a unit watershed and the daily flow monitoring data of an adjacent upstream region. FDC can be prepared with the calculation of daily flow by the equation for each unit watershed. An annual and a full-period FDC were drawn for each unit watershed in Guem river basin. Standard flow such as low and ordinary flow can be obtained from the annual FDC. Major percentile of flow such as 10, 25, 50, 75 or 90% can be obtained from the full-period FDC. It is considered that this simple method of developing FDC can be utilized more widely for the calculation of standard flow and the assessment of water quality in the process of TMDLs.

• Earthquakes and Structures
• /
• 제19권3호
• /
• pp.189-196
• /
• 2020

The near-collapse performance limit is defined as the deformation at the 20% drop of maximum base shear in the decreasing region of the pushover curve for ductile framed buildings. Although monotonic pushover analysis is preferred due to the simple application procedure, this analysis gives rise to overestimated results by neglecting the cumulative damage effects. In the present study, the acceptabilities of monotonic and cyclic pushover analysis results for the near-collapse performance limit state are determined by comparing with Incremental Dynamic Analysis (IDA) results for a 5-story Reinforced Concrete framed building. IDA is performed to obtain the collapse point, and the near-collapse drift ratios for monotonic and cyclic pushover analysis methods are obtained separately. These two alternative drift ratios are compared with the collapse drift ratio. The correlations of the maximum tensile and compression strain at the base columns and beam plastic rotations with interstory drift ratios are acquired using the nonlinear time history analysis results by the simple linear regression analyses. It is seen that these parameters are highly correlated with the interstory drift ratios, and the results reveal that the near-collapse point acquired by monotonic pushover analysis causes unacceptably high tensile and compression strains at the base columns, as well as large plastic rotations at the beams. However, it is shown that the results of cyclic pushover analysis are acceptable for the near-collapse performance limit state.

• 아동학회지
• /
• 제29권5호
• /
• pp.215-226
• /
• 2008

The purpose of this study was to identify dangerous routes by which the experience of victimization leads to bullying behavior and to clarify the intrapersonal moderating variables which control the routes. Subjects were 1,086 students of elementary and middle schools in Seoul and Gyeonggi Province. Data were analyzed by simple regression analysis and multiple moderating regression analysis. The major findings of this study were that (1) the experience of victimization was an important factor predicting bullying behavior; and (2) self esteem and internal locus of control were moderating variables between the experience of victimization and bullying behavior. This study provides effective information to protect students from bullying by finding some moderating variables.

• 한국의류학회지
• /
• 제24권3호
• /
• pp.313-322
• /
• 2000

The purpose of this study was to investigate the characteristic and image of oriental costume designs which represented among three countries, Korea, China and Japan. The specific objectives were: 1) to find out the positioning of oriental costume design. 2) to find out relation to oriental costume image and preference. The stimulus were 75 costume designs of contemporary costume which represented the traditional images of three countries Korea, China and Japan. The main survey of questionary consisted of their evaluation of the oriental costume image by 26 semantic differential bi-polar scales and the subjects were 99 female students majoring in clothing and textile. The data were analyzed by Multidimensional Scaling Method and Regression Analysis. The specific objective were as follows: 1. According to image positioning. The oriental costume design was classified by simple-decorative, soft-hard. 2. As result of regression analysis. The preference of oriental costume image was related to attractive factor.

• Asia pacific journal of information systems
• /
• 제15권2호
• /
• pp.51-63
• /
• 2005

There has been substantial research into the bankruptcy prediction. Many researchers used the statistical method in the problem until the early 1980s. Since the late 1980s, Artificial Intelligence(AI) has been employed in bankruptcy prediction. And many studies have shown that artificial neural network(ANN) achieved better performance than traditional statistical methods. However, despite ANN's superior performance, it has some problems such as overfitting and poor explanatory power. To overcome these limitations, this paper suggests a relatively new machine learning technique, support vector machine(SVM), to bankruptcy prediction. SVM is simple enough to be analyzed mathematically, and leads to high performances in practical applications. The objective of this paper is to examine the feasibility of SVM in bankruptcy prediction by comparing it with ANN, logistic regression, and multivariate discriminant analysis. The experimental results show that SVM provides a promising alternative to bankruptcy prediction.

• Structural Engineering and Mechanics
• /
• 제60권4호
• /
• pp.655-665
• /
• 2016

This paper aims to develop models to accurately predict the behavior of fresh concrete exposed to vibration using artificial neural networks (ANNs) model and regression model (RM). For this purpose, behavior of a full scale precast concrete mold was investigated experimentally and numerically. Experiment was performed under vibration with the use of a computer-based data acquisition system. Transducers were used to measure time-dependent lateral displacements at some points on mold while both mold is empty and full of fresh concrete. Modeling of empty and full mold was made using both ANNs and RM. For the modeling of ANNs: Experimental data were divided randomly into two parts. One of them was used for training of the ANNs and the remaining part was used for testing the ANNs. For the modeling of RM: Sinusoidal regression model equation was determined and the predicted data was compared with measured data. Finally, both models were compared with each other. The comparisons of both models show that the measured and testing results are compatible. Regression analysis is a traditional method that can be used for modeling with simple methods. However, this study also showed that ANN modeling can be used as an alternative method for behavior of fresh concrete exposed to vibration in precast concrete structures.

• 응용통계연구
• /
• 제26권6호
• /
• pp.1009-1018
• /
• 2013

회귀분석은 변수들간의 관계를 파악하는데 유용하게 사용된다. 여러개의 모집단을 비교할 때, 여러 모집단이 갖는 각각의 회귀직선의 기울기가 같은지 검정하는 것이 필요할 때가 있다. 본 논문에서는 순차기울기를 추정한 후 ANOVA의 F-검정법과 Kruskal-Wallis (1952)검정법을 이용한 방법을 각각 제안하였다. 또한, 몬테카를로 모의시험 연구를 통해 본 논문에서 제안한 두 가지 방법과 Park과 Kim (2009)이 제안한 기존 방법의 검정력을 비교하였다.