• Asia pacific journal of information systems
• /
• 제26권1호
• /
• pp.66-79
• /
• 2016

The availability of detailed data on customers' online behaviors and advances in big data analysis techniques enable us to predict consumer behaviors. In the past, researchers have built purchase prediction models by analyzing clickstream data; however, these clickstream-based prediction models have had several limitations. In this study, we propose a new method for purchase prediction that combines information theory with machine learning techniques. Clickstreams from 5,000 panel members and data on their purchases of electronics, fashion, and cosmetics products were analyzed. Clickstreams were summarized using the 'entropy' concept from information theory, while 'random forests' method was applied to build prediction models. The results show that prediction accuracy of this new method ranges from 0.56 to 0.83, which is a significant improvement over values for clickstream-based prediction models presented in the past. The results indicate further that consumers' information search behaviors differ significantly across product categories.

• 수산경영론집
• /
• 제51권1호
• /
• pp.81-96
• /
• 2020

The purpose of this paper is to analyze factors affecting the net income of regional and industrial fisheries cooperatives in South Korea using panel data. This paper utilizes linear or GLS regression models such as pooled OLS model, fixed effects model, and random effects model to estimate affecting factors of the net income of regional and industrial fisheries cooperatives. After reviewing various tests, we eventually select random effects model. The results, based on panel data between 2013 and 2018 year and 64 fisheries cooperatives, indicate that capital and area dummy variables have positive effects and employment has negative effect on the net income of regional and industrial fisheries cooperatives as predicted. However, debt are opposite with our predictions. Specifically, it turns out that debt has positive effect on the net income of regional and industrial fisheries cooperatives although it has been increased. Additionally, this paper shows that the member of confreres does not show any significant effect on the net income of regional and industrial fisheries cooperatives in South Korea. This study is significant in that it analyzes the major factors influencing changes in the net income that have not been conducted recently for the fisheries cooperatives by region and industry.

• 아태비즈니스연구
• /
• 제11권2호
• /
• pp.27-43
• /
• 2020

Purpose - This study explores the impact of the real estate industry on related industries for the perspective of Chinese steel companies. Design/methodology/approach - The impact of housing prices on the 41 listed steel companies' performance was analyzed by using the panel data model. We used two kinds of housing price indexes that are set in the panel data models to estimate the range of the real estate market, driving the performance growth of steel listed companies. Moreover, the net profit of steel companies is used as the dependent variable. To test the stability of the model, ROA used as a dependent variable for the robustness test. Also, to avoid the time trend of housing prices, this paper selects the growth rate of housing prices as the primary research variable. After Fisher-type testings, there is no unit root problem in both independent and dependent variables. Findings - The results indicated that the rise in the housing price has a positive influence on the steel company performance. When the housing price increases by 1%, the net profit of steel enterprises will increase by 5 to 20 million yuan. Research implications or Originality - In this paper, empirical data at the micro-level and panel model are used to quantify China's real estate industry's driving effect on the iron and steel industry, providing evidence from the microdata level. It helps us to understand further the status and role of China's real estate industry in the economic structure.

• International Journal of Naval Architecture and Ocean Engineering
• /
• 제7권2호
• /
• pp.399-408
• /
• 2015

This paper concentrated on the ultimate uniaxial compressive strength of stiffened panel with opening under lateral load and also studied the design-oriented formulae. For this purpose, three series of well executed experiments on longitudinal stiffened panel with rectangular opening subjected to the combined load have been selected as test models. The finite element analysis package, ABAQUS, is used for simulation with considering the large elasticplastic deflection behavior of stiffened panels. The feasibility of the numerical procedure is verified by a good agreement of experimental results and numerical results. More cases studies are executed employing nonlinear finite element method to analyze the influence of design variables on the ultimate strength of stiffened panel with opening under combined pressure. Based on data, two design formulae corresponding to different opening types are fitted, and accuracy of them is illustrated to demonstrate that they could be applied to basic design of practical engineering structure.

• 한국산업보건학회지
• /
• 제34권3호
• /
• pp.279-291
• /
• 2024

Objective: This was to develop headforms for the labor population, based on a three-dimensional(3D) face dimensions data base(DB) and a principal component analysis(PCA) fit test panel, in selection, use and maintenance of respirators in Korea. Methods: This study was part of a two-year-project initiated in 2021. The study was designed and conducted in line with ISO 16976-2 while subjects were those employed in the development of the PCA fit test panel. The approaches included literature review; examination on conformity of the 3D face dimensions DB; and development of headforms representing the labor population. The mean data were used in order to construct each model of the headforms through a way of 3D modeling and 3D printing technology. Results: A total of 2,752 subjects were determined. Five models of headforms(small, medium, large, long-narrow, short-wide) were completely constructed for the labor population. For example, means of the 10 face dimensions for medium headform model were: minimum frontal breadth 106 mm, face width 136 mm, jaw width 127 mm, face length 111 mm, interpupillary distance 69 mm, head breadth 164 mm, nose protrusion 12 mm, nose breadth 34 mm, nasal root breadth 35 mm, and nose length 50 mm. Conclusions: Five models of headforms were newly constructed using the study data. It is likely desirable that the constructed headforms, together with the 3D face dimensions DB as well as the PCA fit test panel, can be utilized more effectively in selection, use and maintenance of respirators for users including the labor population.

• 대한교통학회:학술대회논문집
• /
• 대한교통학회 1998년도 Proceedings The 4th International Transport Symposium
• /
• pp.160-167
• /
• 1998

Recent world-wide interest in activity-based travel behavior modeling has generated an entirely new perspective on how the profession views the travel demand process. This paper seeks to further promote the case of activity-based travel behavior models by providing some empirical evidence of relationship between travel time and activity duration decision for worker with transportation panel data. The travel time from home to work and from work to home, without activity involvement, is estimated by the Ordinary Least Squares (OLS) method. And, the travel time to and from the selected activity and the activity duration are modeled simultaneously by the Three Stage Least Squares (3SLS) method due to the endogenous relationship between travel time and activity duration. Two kinds of models, OLS and 3SLS, include selectivity bias corrections in a discrete/continuous framework, because of the inter-relationship between the choice of activity type/travel mode (discrete) and the travel time/activity duration (continuous). Estimation is undertaken using a sample of over 1300 household two-day trip diaries collected from the same travelers in the Seattle area in 1989. The behavioral consequences of these models provide interesting and provocative findings that should be of value to transportation policy formulation and analysis.

• 부동산연구
• /
• 제27권1호
• /
• pp.33-44
• /
• 2017

최근 중국은 부동산 가격이 급격히 상승하고 있으며, 인플레이션이 부동산 가격의 상승을 유발시킨다는 주장이 제기되고 있다. 따라서 본 연구에서는 2006-2014년의 중국 성별 패널 자료를 이용하여 인플레이션이 중국 부동산시장의 버블에 미치는 영향을 분석하였다. 분석방법은 정태 및 동태(static and dynamic) 패널 데이터 모델을 사용하였다. 본 연구의 분석결과, 인플레이션이 중국 부동산 버블형성의 주요 원인이며 중국 동부지역에서 그 영향이 더욱 뚜렷하게 나타남을 확인할 수 있었다. 우리는 더불어 중국부동산거품이 일정 정도 시장자체조정기제의 영향을 받았다는 흥미로운 현상을 발견했다.

• 한국수자원학회논문집
• /
• 제44권1호
• /
• pp.1-8
• /
• 2011

본 연구는 패널자료(15개 시도의 20년간 자료)를 이용하여 홍수피해 비용과 복구비와의 상관관계에 대한 분석을 시도하였다. 패널 분석은 자료의 성격에 따라 고정효과모형 또는 확률효과모형을 사용하지만 본 분석은 두 모형을 함께 추정하였다. 예상대로 모든 변수들은 복구비와 정의 상관관계를 보였지만 사망자수와 이재민 수는 유의하지 않거나 오히려 음의 상관관계를 보이기도 하였다. 그리고 공공시설의 피해가 가장 중요한 인자였다. 무엇보다 중요한 것은 우리나라의 경우 복구비가 피해액이나 투자액에 비해 항상 높다는 것이다. 우리나라와는 반대로 일본의 경우 복구비가 가장 낮고 투자비가 가장 높다는 점을 미루어 볼 때 결국 적절한 투자가 향후 홍수피해는 물론 복구비도 낮출 수 있다는 점을 확인할 수 있었다.

• 의정연구
• /
• 제23권2호
• /
• pp.147-177
• /
• 2017

한국 선거에서 정치 이념의 역할에 대해 그동안 수많은 연구가 이루어졌으나, 지금까지의 연구들은 예외 없이 횡단면 데이터(cross-sectional data) 분석에 전적으로 의존하였다. 기존의 연구들과 달리 이 연구는 지난 제18대 대통령 선거에서 나타난 이념 투표의 영향을 패널 데이터 분석(panel data analysis)을 통해 보다 엄밀하게 검증한다. 구체적으로, "EAI 총선대선패널조사, 2012"를 이용하여 지난 18대 대통령 선거 시기(2012년 4월 중순-12월 하순)에 정치적이념이 유권자들의 후보 호오도(candidate evaluation)와 후보 선택(vote choice)에 미친 영향을 고정 효과(fixed effects) 분석과 변동 효과(random effect) 분석 등 표준적인 패널 데이터 분석을 통해 살펴보았다. 또한 이념투표의 대표적 설명 모형인 근접성 모형(proximity model)과 방향성 모형(directional change model)을 적용한 결과를 비교 분석하였다. 분석 결과, 지난 18대 대통령 선거에서 진보와 보수의 정치적 이념은 박근혜와 문재인후보에 대한 호오도 및 이들에 대한 지지에 독립적이고 유의미한 영향을 미쳤다고 할 수 있다. 즉, 유권자들은 특정 후보와의 이념적 차이가 크다고 느낄수록 그후보에 대한 지지 확률 및 호오도가 상대적으로 낮아졌다. 그리고 이러한 이념투표의 영향은 근접성 모형보다 방향성 모형에서 상대적으로 뚜렷하게 나타났다.

• 반도체디스플레이기술학회지
• /
• 제11권1호
• /
• pp.67-71
• /
• 2012

The optical properties of PDP, such as the transmittance and reflectance, were analyzed with 3D optical code. Three different ITO-less structures in the front panel are examined. In the assembled panel study, the test 1 structure shows 16.6% and 10.2% higher reflectance than the structures of tests 2 and 3, respectively. In order to check the validation of the simulation result, three 7.5-inch test panels having the same geometry and property are fabricated as simulation models. The calculated reflective properties are compared to the measured data from real panels. The relative difference extracted from the simulation and measurement methods is less than 4.9% and are well matched.