• The Korean Journal of Applied Statistics
• /
• v.33 no.1
• /
• pp.1-10
• /
• 2020

We propose a unified approach to variable selection, transformation and outliers in the linear model. The procedure includes a sequential method for outlier detection and a least trimmed squares estimator for variable transformation. It uses all possible subsets regressions for model selection. Some real data analyses and the simulation results are provided to show the efficiency of the methods in the context of the correct variable selection and the fitness of the estimated model.

• The KIPS Transactions:PartD
• /
• v.16D no.6
• /
• pp.945-960
• /
• 2009

The purpose of this study is to define determinants of VoIP user acceptance and to verify significant causality among latent variables - performance expectancy, effort expectancy, cost expectancy, social influence, facilitating conditions, behavioral intend, use behavior - based on UTAUT model. We presented the expanded hypotheses including the new factor, cost expectancy and analyzed the moderating effect of user age, gender and usage experience variables. For a accuracy of predicted results, we focused on survey analysis with 641 real user samples. Compared to previous studies, it is meaningful that this research verified the conceptual difference between behavioral intention and usage behavior. As a result, all proposed hypotheses accepted and moderating effects are supported significantly in age and use experience moderating variables.

• The Korean Journal of Applied Statistics
• /
• v.13 no.1
• /
• pp.189-196
• /
• 2000

Conjugate prior density families were motivated by considerations of tractability in implementing the Bayesian paradigm. But we consider problem that the conjugate prior p($\Theta$) cannot be used in restriction of the parameter $\Theta$. This article considers the nonconjugate prior problem of hierarchical Poisson model. We demonstrate the use of latent variables for sampling non-standard densities which arise in the context of the Bayesian analysis of non-conjugate by using a Gibbs sampler.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2011.05a
• /
• pp.224-224
• /
• 2011

5대 주요 수계의 소수력자원에 대하여 연구를 수행하였고, 하천의 유량지속특성을 예측할 수 있는 모델이 개발되었으며, 이를 이용하면 강우사상으로 야기되는 유입량의 변화에 대한 분석이 가능하다. 또한 소수력발전소의 성능을 예측할 수 있는 모델도 개발되었다. 안동댐에서 측정된 월유입량 자료를 분석하였으며,. 본 연구를 통해 개발된 모델을 이용하여 예측한 결과는 안동댐에서 오랜기간 동안 측정된 결과와 거의 일치되는 것으로 나타났다. 이는 개발된 모델들이 소수력발전입지의 이용가능한 잠재량과 기술적 잠재량을 예측하는데 효과적인 것으로 밝혀졌다. 본 모델들을 이용하여 수계별로 소수력발전입지에 대한 수문학적 성능을 분석하였다. 분석결과 소수력발전 입지의 수문학적 성능특성은 수계별로 차이가 있는 것으로 나타났다. 특히 북한강과 낙동강수계에 위치한 소수력발전입지의 비설계유량과 비출력량은 다른 수계들 보다 차이가 큰 것으로 나타났다. 그림 1은 수계별 비설계유량에 따른 비출력량의 변화를 나타낸다.

• Journal of Global Scholars of Marketing Science
• /
• v.15 no.2
• /
• pp.141-159
• /
• 2005

In this research, we develope an estimation method for the estimation of the market potential in the new service (or product) diffusion model. The developed method is based on the ordered response model which can effectively incorporate the survey result of the multi-point scale intention for subscription as well as the responder's characteristics, the characteristics & attitudes of the related service. We also apply the developed method to an estimation of the market potential of the digital multimedia broadcasting (DMB) service. As a result, an optimistic and a pessimistic estimates of DMB market potential are 41.10% and 14.83% of the cellular subscribers respectively.

• Journal of the Korean Data and Information Science Society
• /
• v.25 no.2
• /
• pp.411-422
• /
• 2014

This study analyzed how average GPA (grade point average) changes as the number of completed semesters increases based on the estimates of intercept, slope, and quadratic term. The students included in this study are those who was admitted in 2011 and took 6 consecutive semesters. More precisely, it was analyzed if intercept, slope and quadratic term of average GPA were different between gender and selection method. The results showed that the intercept was different between selection method, the slope was different between gender, but the quadratic term was different between neither selection method nor gender.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2021.06a
• /
• pp.115-115
• /
• 2021

식생은 증발산, 강우, 토양 수분 등 다양한 수문기상 요인과 밀접한 관계가 있기 때문에 식생의 상태는 가뭄 발생 시 물 부족에 매우 큰 영향을 받는다. 가뭄에 따른 식생의 변화와 영향을 파악하기 위해서는 식생-기후의 피드백을 이해해야 한다. 식생과 기후변수의 상호관계를 묘사하고 결합 확률을 구성하는 것은 식생-기후의 피드백을 이해하는데 적절하다. Copula 함수는 모든 변수를 연결하는 이점을 가지기 때문에 다양한 확률 변수를 결합하는 강력한 접근방법으로, copula를 통한 확률론적 접근방법은 수문 기상 스트레스에 대한 식생의 반응을 효과적으로 조사할 수 있다. 이에 따라 본 연구에서는 copula 기반의 식생-기후의 상호관계를 통해 가뭄 발생 시 식생이 받을 수 있는 영향을 정량화하고자 한다. 이를 위해 위성 자료를 활용한 식생건강성지수(Vegetation Health Index, VHI)와 위성관측된 강수 및 잠재증발산 자료를 적용하여 높은 공간 해상도에서 한국 전역의 식생 가뭄 가능성을 추정하고자 하였다. 강수 및 잠재증발산 자료를 통해 다양한 가뭄지수를 산정하고, copula 결합 이론을 기반으로 VHI와 가뭄지수 간의 이변량 결합 확률모델이 제안된다. 이에 조건부 확률을 적용하여 다양한 가뭄 시나리오에서 식생의 가뭄 가능성을 추정하고, 가뭄에 취약한 지역을 공간적으로 분석하고자 한다. 이를 통해 가뭄 스트레스에 따른 식생 변화와 생태학적 가뭄의 공간적 특성을 효과적으로 파악할 수 있을 것으로 기대된다.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.22 no.6
• /
• pp.49-65
• /
• 2023

In response to the growing global concern for the environment, the international community has recently committed to achieving 'carbon neutrality.' As a result, numerous studies have been conducted on mode choice models that include carbon emissions as a variable. However, few studies have established a correlation between individual preferences and carbon emissions. In this study, a new mode of transportation named sustainable public transit (SPT), incorporating carbon-reducing transport options like electric scooters, is proposed. Analyzing the individual preferences of commuters on carbon emissions through factor analysis, a stated preference (SP) survey was conducted. A mode choice model for SPT was constructed using multinomial logit models. The results of the analysis showed that gender, income, and specific preferences, such as a passion for exploring new routes, a preference for intermodal transfers, knowledge of carbon reduction, and carbon reduction practices, significantly influence latent preferences for SPT. Therefore, this study is significant as it considers carbon emissions as an attribute variable during the construction of mode choice models and reflects the individual preference variables associated with carbon reduction.

• Asia-Pacific Journal of Business Venturing and Entrepreneurship
• /
• v.16 no.1
• /
• pp.37-53
• /
• 2021

The purpose of this study is to analyze the types of latent profiles depending on the job satisfaction of youth entrepreneurs in South Korea and examine the influence of related variables. Latent profile analysis (LPA) was conducted using data from the 12th year of the Youth Panel Study. Results of the study revealed that there were five latent profiles based on job satisfaction among young entrepreneurs: lowest job satisfaction, mid-low job satisfaction, middle job satisfaction, mid-high job satisfaction, and highest job satisfaction. By applying multinomial logistic analysis, the influencing factors for each latent profiles have been derived. Gender, marital status, subjective health, self-esteem, entrepreneurial motivation, difficulty in start-up, person-job fit, intrinsic or extrinsic job values, business scale, income, work hour, education or training, and labor insurance were significant predictors of youth entrepreneurs' job satisfaction. Based on these results, the article discusses practical implications that can enhance the job satisfaction of young entrepreneurs.

• The Journal of the Acoustical Society of Korea
• /
• v.41 no.6
• /
• pp.662-667
• /
• 2022

Computing music similarity is an indispensable component in developing music search service. This paper focuses on learning a music similarity function in order to boost cover song identification performance. By using the probabilistic linear discriminant analysis, we construct a latent music space where the distances between cover song pairs reduces while the distances between the non-cover song pairs increases. We derive a music similarity function by testing hypothesis, whether two songs share the same latent variable or not, using the probabilistic models with the assumption that observed music features are generated from the learned latent music space. Experimental results performed on two cover music datasets show that the proposed music similarity improves the cover song identification performance.