• The Korean Journal of Applied Statistics
• /
• v.14 no.2
• /
• pp.475-486
• /
• 2001

데이터마이닝 패키지에 구현된 분류나무 알고리즘 가운데 CART, CHAID, QUEST, C4.5에서 변수 선택법을 비교하였다. CART의 전체탐색법이 편의를 갖는다는 사실은 잘알려졌으며, 여기서는 상품화된 패키지들에서 이들 알고리즘의 편의와 선택력을 모의실험 연구를 통하여 비교하였다. 상용 패키지로는 CART, Enterprise Miner, AnswerTree, Clementine을 사용하였다. 본 논문의 제한된 모의실험 연구 결과에 의하면 C4.5와 CART는 모두 변수선택에서 심각한 편의를 갖고 있으며, CHAID와 QUEST는 비교적 안정된 결과를 보여주고 있었다.

• Proceedings of the Korean Statistical Society Conference
• /
• 2003.10a
• /
• pp.263-268
• /
• 2003

Breiman, Friedman, Olshen and Stone(1984)의 전체탐색법에 의한 회귀나무는 상대적으로 많은 분리가 가능한 변수로 분리기준이 정해지는 편의 현상을 갖고 있다. 본 연구에서는 이런 문제점을 해결할 수 있는 알고리즘을 제안하여 변수선택편의가 없는 회귀나무를 만들고자 한다. 제안하는 알고리즘은 노드의 분리변수를 선택하는 단계와 그 선택된 변수에 의해 이진분리를 위한 분리점을 찾는 단계로 구성되어 있다. 예측변수 중에서 목표변수와 가장 밀접하게 연관된 예측변수는 예측변수의 자료의 종류에 따라 스피어만의 순위상관계수에 의한 검정 혹은 크루스칼-왈리스의 통계량에 의한 검정을 수행하여 가장 통계적으로 유의한 변수로 선택하였고, 선택된 변수에만 Breiman et al.(1984)의 전체선택법을 적용하여 분리점을 결정하였다. 모의실험을 통해 변수선택편의, 변수선택력 , 그리고 평균제곱오차 측면에서 Breiman et al. (1984)의 CART(Classification and Regression Trees)와 제안한 알고리즘을 서로 비교하였다. 또한, 두 알고리즘을 실제 자료에 적용하여 효율을 서로 비교하였다.

• Journal of Labour Economics
• /
• v.42 no.3
• /
• pp.39-74
• /
• 2019

In this paper, the wage equations of local labor markets for college graduates in Korea are estimated by Dahl(2002)'s methodology to correct for selection bias. The results suggest that the variations of coefficients in wage equations across the local labor markets are mostly remained after correcting for selection bias. The gender wage gap is hardly affected by selection bias. The variations of return to education and the major premium are reduced about 18% and 11% respectively. Meanwhile, the selection bias is negligible in the national capital region, which suggests that college graduates prefer the national capital region regardless of their gender, level of education, and major.

• The Korean Journal of Applied Statistics
• /
• v.29 no.2
• /
• pp.389-398
• /
• 2016

In this paper, we examine the financial performance of credit guarantee programs. We compared financial performance of guaranteed firms of KODIT and non-guaranteed firms. The of covariate adjusted propensity score method is used because a selection bias problem could occur if t-test or regression analysis were used. The results show that a credit guarantee program enhances the financial performance of beneficiary firms.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.52 no.10
• /
• pp.73-81
• /
• 2015

The paper proposes a prototype selection method and evaluates the generalization performance of standard algorithms and prototype based classification learning. The proposed prototype classifier defines multidimensional spheres with variable radii within class areas and generates a small set of training data. The nearest-neighbor classifier uses the new training set for predicting the class of test data. By decomposing bias and variance of the mean expected error value, we compare the generalization errors of k-nearest neighbor, Bayesian classifier, prototype selection using fixed radius and the proposed prototype selection method. In experiments, the bias-variance changing trends of the proposed prototype classifier are similar to those of nearest neighbor classifiers with all training data and the prototype selection rates are under 27.0% on average.

• The Korean Journal of Applied Statistics
• /
• v.17 no.2
• /
• pp.347-357
• /
• 2004

It is very important to select a split variable in constructing the classification tree. The efficiency of a classification tree algorithm can be evaluated by the variable selection bias and the variable selection power. The C4.5 has largely biased variable selection due to the influence of many distinct values in variable selection and the QUEST has low variable selection power when a continuous predictor variable doesn't deviate from normal distribution. In this thesis, we propose the SRT algorithm which overcomes the drawback of the C4.5 and the QUEST. Simulations were performed to compare the SRT with the C4.5 and the QUEST. As a result, the SRT is characterized with low biased variable selection and robust variable selection power.

• Culinary science and hospitality research
• /
• v.20 no.6
• /
• pp.147-158
• /
• 2014

The purpose of this study is to investigate the preference and present use condition of convenience foods for North Korean refugees in South Korea. Questionnaires were completed by 211 North Korean refugees and data was analyzed with SPSS software. The preference for convenience foods of half of the questionnaires was investigated as 'moderate'. More than half of the North Korean refugees enjoy eating the convenience foods. They enjoy them as a snack. The reason to eat convenience foods is due to the convenience(n=122, 57.8%), and taste(n=42, 10.0%). North Korean refugees under the age of 29 years think the fastfood is good due to the convenience and taste. The longer the period of residence in South Korea, the less they enjoy convenience foods. The reason to choose the convenience foods is 'good for health', 'convenient', 'exotic', and 'economical'. The results suggest that it is necessary to educate people to buy reasonably by understanding the relationship between the convenience foods and health.

• Journal of the Korean Data and Information Science Society
• /
• v.27 no.5
• /
• pp.1317-1325
• /
• 2016

The purpose of this paper is to explain the factors affecting the wage of the vocational high school graduates. We particularly examine the effectiveness of controlling sample selection bias by employing the Tobit model and Heckman sample selection model. The major results are as follows. First it is shown that the Tobit model and Heckman sample selection model controlling sample selection bias is statistically significant. Hence all the independent variables seem to be statistically consistent with the theoretical model. Second, gender was statistically significant, both in the probability of employment and the wage. Third, the employment probability and wage of Maester high school graduates were shown to be high compared to all other graduates. Fourth, the higher parent's income, the higher are both the employment probability and the wage. Finally, parents education level, high school grade, satisfaction, and a number of licenses were found to be statistically significant, both in the probability of employment and wages.

• The Korean Journal of Applied Statistics
• /
• v.29 no.4
• /
• pp.643-656
• /
• 2016

High dimensional data are frequently encountered in various fields where the number of variables is greater than the number of samples. It is usually necessary to select variables to estimate regression coefficients and avoid overfitting in high dimensional data. A penalized regression model simultaneously obtains variable selection and estimation of coefficients which makes them frequently used for high dimensional data. However, the penalized regression model also needs to select the optimal model by choosing a tuning parameter based on the model selection criterion. This study deals with the bias effect of LASSO regression for model selection criteria. We numerically describes the bias effect to the model selection criteria and apply the proposed correction to the identification of biomarkers for lung cancer based on gene expression data.

• Journal of Korean Society of Transportation
• /
• v.20 no.3
• /
• pp.83-92
• /
• 2002

Freight mode choice models are essential to the analysis of many areas of transport research. However, observations of actual market choices have only been made in a limited number of situations. Therefore, stated preference(SP) techniques have emerged as an alternative source of actual market choices to be used for estimating freight mode choice models. Considerable confidence exists about SP data, but little consideration has been given to the potential for estimation bias. This paper has been motivated by the theoretical side of estimating SP discrete choice models, focusing on a case study of freight mode choice. Recently developed simulation methods are used to construct inherent random heterogeneity legit models, which consider individual heterogeneity, its inheritance to the next choices and overcome the independence from irrelevant alternatives (IIA) property. This Paper contributes to the development of models dealing with heterogeneity and its inheritance, and sheds light on the heterogeneity of freight transport.