• Analytical Science and Technology
• /
• v.28 no.3
• /
• pp.187-195
• /
• 2015

The heat of sublimation (HOS) is an essential parameter used to resolve environmental problems in the transfer of organic contaminants to the atmosphere and to assess the risk of toxic chemicals. The experimental measurement of the heat of sublimation is time-consuming, expensive, and complicated. In this study, quantitative structural property relationships (QSPR) were used to develop a simple and predictive model for measuring the heat of sublimation of organic compounds. The population-based forward selection method was applied to select an informative subset of descriptors of learning algorithms, such as by using multiple linear regression (MLR) and the support vector machine (SVM) method. Each individual model and consensus model was evaluated by internal validation using the bootstrap method and y-randomization. The predictions of the performance of the external test set were improved by considering their applicability to the domain. Based on the results of the MLR model, we showed that the heat of sublimation was related to dispersion, H-bond, electrostatic forces, and the dipole-dipole interaction between inter-molecules.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.14 no.1
• /
• pp.14-21
• /
• 2021

Green networking has become a issue to reduce energy wastes and CO2 emission by adding energy managing mechanism to wired data networks. Energy consumption of the overall wired data networks is driven by access networks, expect for end devices. However, on a global scale, it is more difficult to manage centrally energy, measure and model the real energy use and energy savings potential of the access networks. This paper presented the multiple linear regression model to predict energy consumption of wired access networks using supervised learning of machine learning with data collected by existing investigated materials, actual measured values and results of many models. In addition, this work optimized the performance of it by various experiments and predict energy consumption of wired access networks. The performance evaluation of the regression model was achieved by well-knowned evaluation metrics.

• The Journal of the Acoustical Society of Korea
• /
• v.4 no.3
• /
• pp.42-46
• /
• 1985

자동차에 의해 발생되는 도로교통소음을 예측할 수 있는 기본모델을 개발하고저 한다. 이를 위 해 기존도로중에서 통행방법 등을 조사하고저 통행량, 속도등의 통행방법 및 측정거리가 조사되었고, 소 음평가량으로서 등가소음수준 Leq와 소음수준 중앙치 L\sub 50\도 측정되었다. 본 연구에서는 이와 같 은 자료를 토대로 측정된 자료를 토대로 선형회귀분석 방법을 사용한다. 이렇게 개발된 모델을 동일한 조건에서 실측된 자료에 적용한 결과 정확도가 상당히 높았다. 다른 지역에서 이미 개발된 모델로서는 수학적인 모델과 통계적인 모델들이 있다. 이미 개발된 모델들과는 실측치와 예측치와의 오차의 제곱을 합계한 값으로서 비교했다.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2003.05a
• /
• pp.85-88
• /
• 2003

강화학습의 한가지 방법인 Q-learning은 최근에 Linear Quadratic Regulation(이하 LQR) 문제에 성공적으로 적용된 바 있다. 특히, 시스템 모델의 파라미터에 대한 구체적인 정보없이 적절한 입ㆍ출력만으로 학습을 통해 문제의 해결이 가능하므로 상황에 따라 매우 실용적인 방법이 될 수 있다. 뉴럴-큐 기법은 이러한 Q-learning의 Q-value를 MLP(multilayer perceptron) 신경망의 출력으로 대치시켜, 비선형 시스템의 최적제어 문제를 다룰 수 있게 한 방법이다. 그러나, 뉴럴-큐 기법은 신경망의 구조를 먼저 결정한 후 역전파 알고리즘을 이용해 학습하는 절차를 행하므로, 시행착오를 통해 신경망 구조를 결정해야 한다는 점, 역전파 알고리즘의 적용에 따라 신경망의 연결강도 값들이 지역적 최적해로 수렴한다는 점등의 문제점이 있다. 본 논문에서는 뉴럴-큐 학습의 도구로 KFD회귀를 이용하여 Q 함수의 근사 기법을 제안하고 관련 수식을 유도하였다. 그리고, 모의 실험을 통하여, 제안된 뉴럴-큐 방법의 적용 가능성을 알아보았다.

• Journal of Industrial Convergence
• /
• v.16 no.4
• /
• pp.33-46
• /
• 2018

Personal credit scoring is an effective tool for banks to properly guide decision profitably on granting loans. Recently, many classification algorithms and models are used in personal credit scoring. Personal credit scoring technology is usually divided into statistical method and non-statistical method. Statistical method includes linear regression, discriminate analysis, logistic regression, and decision tree, etc. Non-statistical method includes linear programming, neural network, genetic algorithm and support vector machine, etc. But for the development of the credit scoring model, there is no consistent conclusion to be drawn regarding which method is the best. In this paper, we will compare the performance of the most common scoring techniques such as logistic regression, neural network, and support vector machines using personal credit data of the financial institution in China. Specifically, we build three models respectively, classify the customers and compare analysis results. According to the results, support vector machine has better performance than logistic regression and neural networks.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2000.04a
• /
• pp.227-234
• /
• 2000

지금까지 수행된 대부분의 유가예측은 주고 계량 데이터를 기반으로 하는 여러 가지 계량 모델을 구성하여 수행되었으며, 그 결과 산유국 동향과 같은 국제 유가시장의 불확실성을 제대로 반영하지 못했다. 따라서, 본 논문에서는 이러한 문제점을 해결하기 위하여 계량경제학적인 접근방법과 전문가시스템을 통합한 유가예측 시스템을 설계 및 구현하였다. 즉, 계량 데이터를 기초로 유가예측 모델을 구성하고, 산유국동향과 같은 비계량적인 요인이 유가에 미치는 영향에 대한 실무자의 경험적인 지식은 지식베이스로 구축함으로써, 유가예측과 관련된 다양한 요인들을 폭넓게 고려할 수 있는 통합된 시스템을 개발하였다. 유가예측 모델로는 대표 유종의 유가 및 수급 전망을 위한 동적 선형연립 모델과 유종간 유가의 균형차액을 예측하기 위한 Fully Modified 공적분 회귀분석 모델을 구성하였으며, 유가예측 모델에서 반영하기 어려운 산유국 동향이나 OPEC정책, 선물시장 동향 등은 실무자의 경험적인 지식을 바탕으로 시스템 예측변수로 설정하여 유가예측에 반영할 수 있도록 지식베이스를 구축하였다. 또한, 본 시스템에서는 유가예측 이외에 석유 수급을 전망하고, 유가 및 수급과 관련된 다양한 정보를 제공하고 관리하는 기능을 제공하고 있다.

• Korean Journal of Construction Engineering and Management
• /
• v.20 no.6
• /
• pp.126-131
• /
• 2019

The estimation of available solar energy at particular locations is critical to find and assess suitable locations of PV sites. The amount of PV power generation is however affected by various geographical factors (e.g., weather), which may make it difficult to identify the complex relationship between affecting factors and power outputs and to apply findings from one study to another in different locations. This study thus undertakes a regression analysis using data collected from 172 PV plants spatially distributed in Korea to identify critical weather conditions and estimate the potential power generation of PV systems. Such data also include solar radiation, precipitation, fine dust, humidity, temperature, cloud amount, sunshine duration, and wind speed. The estimated PV power generation is then compared to the actual PV power generation to evaluate prediction performance. As a result, the proposed model achieves a MAPE of 11.696(%) and an R-squred of 0.979. It is also found that the variables, excluding humidity, are all statistically significant in predicting the efficiency of PV power generation. According, this study may facilitate the understanding of what weather conditions can be considered and the estimation of PV power generation for evaluating and determining suitable locations of PV facilities.

• International Journal of Highway Engineering
• /
• v.9 no.2 s.32
• /
• pp.1-11
• /
• 2007

The rapid growths of economy and automobiles since the 1970's have caused serious traffic jams and environmental disruption in urban areas. To relieve these problems caused by urbanization, there should be considered alternative means of transportation modes. Many developed countries have accepted bicycles as a so called "Green Mode" for environmentally oriented strategies to increase the qualities of urban lives. Korea have also attempted various means to raise bicycle usages. In this research, significant factors affecting bicycle crashes at signalized intersections in urban areas were studied. The model results showed that Poisson regression is the best fit methodology for data modeling and revealed that traffic volume, a number of driveways, configuration of the ground, presence of bicycle path, school, and bus stop, residential area, size of intersection are significant factors affecting the bicycle crashes.

• Journal of Korean Society of Transportation
• /
• v.23 no.1
• /
• pp.33-44
• /
• 2005

Railroad crossing crashes are fewer than road crashes, but with regard to crash severity, they can be serious injury crashes. There should be, therefore, enormous efforts to increase the safety of railroad crossings. The objective of this paper is to identify and understand factors associated with railroad crossing crashes. Statistical models are used to examine the relationships between crossing accidents and geometric elements of crossings. The results show the Poisson model is the most appropriate method for the crossing accidents, because overdispersion was not observed. This study identifies seven significant factors associated with railroad crossing crashes through the main and variant models. With regard to explanatory factors on crossing safety, the total traffic volume, daily train volume, presence of commercial area around crossings, distance of train detector from crossings, time duration between the activation of warning signals and gates, crossing types, and speed hump were found to affect the safety of railroad crossings.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.29 no.4A
• /
• pp.267-279
• /
• 2009

A numerical model that can simulate the nonlinear behavior of ultra high strength fiber reinforced concrete (UHSFRC) structures subjected to monotonic loading is introduced. The material properties of UHSFRC, such as compressive and tensile strength or elastic modulus, are different from normal strength reinforced concrete. The uniaxial compressive stress-strain relationship of UHSFRC is designed on the basis of experimental result, and the equivalent uniaxial stress-strain relationship is introduced for proper estimation of UHSFRC structures. The steel is uniformly distributed over the concrete matrix with particular orientation angle. In advance, this paper introduces a numerical model that can simulate the tension-stiffening behavior of tension part of the axial member on the basis of the bond-slip relationship. The reaction of steel fiber is considered for the numerical model after cracks of the concrete matrix with steel fibers are formed. Finally, the introduced numerical model is validated by comparison with test results for idealized UHSFRC beams.