• Proceedings of the KIEE Conference
• /
• 2001.11b
• /
• pp.432-434
• /
• 2001

This paper proposes a special-day load forecasting method with the characteristics of temperature based on fuzzy linear regression. We can obtain a linear regression model from the relation between daily peak load and daily maximum or minimum temperature. Simulation results show that the proposed method can improve an accuracy of a special-day load forecasting.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2002.05a
• /
• pp.792-796
• /
• 2002

관찰적 작업자세 평가기법은 각 관절의 자세를 관찰 기록하여 자세 부하를 평가하는 실용적인 인간공학적 작업평가 기법이다. 본 연구에서는 각 관절의 불편도 지수와 전신의 자세 부하의 관계성을 모형화하고, 전신의 작업자세 부하를 평가하는 방법론을 제시하였다. 자동차 조립공정의 대표적인 작업자세들을 대상으로 하여 정적인 자세의 심물리학적 부하를 전신에 대하여 평가하였다 전신의 불편도는 비중립 자세를 취하고 있는 각 관절의 조합에 의해 영향을 받는다. 특히, 자동차 조립공정에서는 어깨 높이 이상의 작업을 대상으로 하는 경우에 어깨, 목, 허리, 손목 등에서 비중립 자세를 동시에 취하여 전신의 불편도가 큰 것으로 나타났다. 평가된 전신의 불편도와 각 자세의 관절별 불편도 지수의 관계를 다중선형회귀모형으로 모형화하는 것이 타당한 것으로 나타났다. 모형에서 전신 불편도에 가장 큰 영향을 미치는 관절은 어깨이며, 손목의 영향이 가장 적은 것으로 나타났다. 이 모형을 통해 작업자세 부하를 정량적으로 평가하는 것이 가능할 것으로 기대된다.

• Computational Structural Engineering
• /
• v.10 no.2
• /
• pp.265-272
• /
• 1997

The purpose of this paper is to develop Neural Network models for Approximate Structural Analysis (NNASA). As an initial stage, the paper classifies the characteristics and the active role of neural networks in the numerical analysis by comparing neural networks with conventional numerical analysis algorithms. The paper proposed two methods of finding solutions of linear algebraic equations by a modified neural network algorithm, and presents that multilayer feedforward networks are a class of universal approximators by comparing the neural network with regression and interpolation techniques.

• 기계와재료
• /
• v.22 no.1
• /
• pp.46-53
• /
• 2010

공작기계에서 가공정밀도를 저하시키는 가장 큰 요인은 열변형 및 채터진동이다. 본 고에서는 이 중 장시간 가공중 기계의 열변형에 따른 문제점을 자동으로 공작기계 CNC(Computerized Numerical Controller) 제어기상에서 실시간으로 보상하여 주는 장치 및 기술개발 사례에 대한 내용을 언급하고자 한다. 기계가공에서 온도신호의 실시간 데이터 취득 및 열변형에 따른 공작기계 원점(Work Offset)의 자율보정이 가공정밀도 향상 및 가동률 향상에 많은 영향을 끼친다 이에 따라 본 고에서는 온도 데이터의 취득부와 보상을 위한 보정값 추출을 위한 선형회귀법 및 신경회로망의 보정모델을 임베디드화한 디바이스와 CNC상에서 가공중 공작기계 원점 자동보정을 하는 시스템을 개발하였기에 관련내용을 소개하고자 한다.

• Journal of the Semiconductor & Display Technology
• /
• v.19 no.4
• /
• pp.59-64
• /
• 2020

In this paper, we found the solution using data based machine learning regression method to check the pore shape, to solve the problem of the experiment quantity occurring when producing scaffold with the 3d printer. Through experiments, we learned secured each print condition and pore shape. We have produced the scaffold from scaffold pore shape defect prediction model using multiple linear regression method. We predicted scaffold pore shapes of unsecured print condition using the manufactured scaffold pore shape defect prediction model. We randomly selected 20 print conditions from various predicted print conditions. We print scaffold five times under same print condition. We measured the pore shape of scaffold. We compared printed average pore shape with predicted pore shape. We have confirmed the prediction model precision is 99 %.

• Journal of Korean Society of Steel Construction
• /
• v.20 no.1
• /
• pp.21-31
• /
• 2008

response surface method (RSM) is widely used to evaluate th e extremely smal probability of ocurence or toanalyze the reliability of very complicated structures. Althoug h Monte-Carlo Simulation (MCS) technique can evaluate any system, the procesing time of MCS dependson the reciprocal num ber of the probability of failure. The stochastic finite element method could solve thislimitation. However, it is limit ed to the specific program, in which the mean and coeficient o f random variables are programed by a perturbation or by a weigh ted integral method. Therefore, it is not aplicable when erequisite programing. In a few number of stage analyses, RSM can construct a regresion model from the response of the c omplicated structural system, thus, saving time and efort significantly. However, the acuracy of RSM depends on the dist ance of the axial points and on the linearity of the limit stat e functions. To improve the convergence in exact solution regardl es of the linearity limit of state functions, an improved adaptive response surface method is developed. The analyzed res ults have ben verified using linear and quadratic forms of response surface functions in two examples. As a result, the be st combination of the improved RSM techniques is determined and programed in a numerical code. The developed linear adapti ve weighted response surface method (LAW-RSM) shows the closest converged reliability indices, compared with quadratic form or non-adaptive or non-weighted RSMs.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.1
• /
• pp.20-26
• /
• 2019

Due to the human influence of particulate matter, various studies are being conducted to predict it using past data measured in the atmospheric environment monitoring network. However, it is difficult to precisely set the measurement environment and detailed conditions of the previously designed predictive model, and it is necessary to design a new predictive model based on the existing research results because of the problems such as the missing of the weather data. In this paper, as a previous study for particulate matter prediction, the conformity of the algorithm for particulate matter prediction was evaluated by designing the prediction model through the multiple linear regression and the artificial neural network, which are machine learning algorithms. As a result of the prediction performance comparison through RMSE, 18.13 for the MLR model and 14.31 for the MLP model, and the artificial neural network model was more conformable for predicting the particulate matter concentration.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.27 no.1
• /
• pp.13-21
• /
• 2021

The purpose of this study is to develop the ship valuation model by utilizing the neural network model. The target of the valuation was secondhand VLCC. The variables were set as major factors inducing changes in the value of ship through prior research, and the corresponding data were collected on a monthly basis from January 2000 to August 2020. To determine the stability of subsequent variables, a multi-collinearity test was carried out and finally the research structure was designed by selecting six independent variables and one dependent variable. Based on this structure, a total of nine simulation models were designed using linear regression, neural network regression, and random forest algorithm. In addition, the accuracy of the evaluation results are improved through comparative verification between each model. As a result of the evaluation, it was found that the most accurate when the neural network regression model, which consist of a hidden layer composed of two layers, was simulated through comparison with actual VLCC values. The possible implications of this study first, creative research in terms of applying neural network model to ship valuation; this deviates from the existing formalized evaluation techniques. Second, the objectivity of research results was enhanced from a dynamic perspective by analyzing and predicting the factors of changes in the shipping. market.

• The Transactions of the Korea Information Processing Society
• /
• v.13 no.6
• /
• pp.260-268
• /
• 2024

Cervical cancer is the fourth most common cancer in women worldwide, and more than 604,000 new cases were reported in 2020 alone, resulting in approximately 341,831 deaths. The Cox regression model is a major model widely adopted in cancer research, but considering the existence of nonlinear associations, it faces limitations due to linear assumptions. To address this problem, this paper proposes ResSurvNet, a new model that improves the accuracy of cervical cancer mortality prediction using ResNet's residual learning framework. This model showed accuracy that outperforms the DNN, CPH, CoxLasso, Cox Gradient Boost, and RSF models compared in this study. As this model showed accuracy that outperformed the DNN, CPH, CoxLasso, Cox Gradient Boost, and RSF models compared in this study, this excellent predictive performance demonstrates great value in early diagnosis and treatment strategy establishment in the management of cervical cancer patients and represents significant progress in the field of survival analysis.

• Journal of KIISE:Software and Applications
• /
• v.33 no.2
• /
• pp.252-266
• /
• 2006

Many problems in computer vision are mainly based on mathematical models. Their optimal solutions can be found by estimating the parameters of each model. However, provided an input data set is involved outliers which are relative]V larger than normal noises, they lead to incorrect results. RANSAC is a representative robust algorithm which is used to resolve the problem. One major problem with RANSAC is that it needs priori knowledge(i.e. a percentage of outliers) of the distribution of data. To solve this problem, we propose a FRANSAC algorithm which improves the rejection rate of outliers and the accuracy of solutions. This is peformed by categorizing all data into good sample set, bad sample set and vague sample set using a fuzzy classification at each iteration and sampling in only good sample set. In the experimental results, we show that the performance of the proposed algorithm when it is applied to the linear regression and the calculation of a homography.