• Journal of the Earthquake Engineering Society of Korea
• /
• v.20 no.1
• /
• pp.45-54
• /
• 2016

This study suggests a prediction model of ground motion spectral shape considering characteristics of earthquake records in Korea. Based on the Graizer and Kalkan's prediction procedure, a spectral shape model is defined as a continuous function of period in order to improve the complex problems of the conventional models. The approximate spectral shape function is then developed with parameters such as moment magnitude, fault distance, and average shear velocity of independent variables. This paper finally determines estimator coefficients of subfunctions which explain the corelation among the independent variables using the nonlinear optimization. As a result of generating the prediction model of ground motion spectral shape, the ground motion spectral shape well estimates the response spectrum of earthquake recordings in Korea.

• The Transactions of the Korean Institute of Electrical Engineers
• /
• v.45 no.4
• /
• pp.602-611
• /
• 1996

In this paper, radial basis function networks with two hidden layers, which employ the K-means clustering method and the hierarchical training, are proposed for improving the short-term predictability of chaotic time series. Furthermore the recursive training method of radial basis function network using the recursive modified Gram-Schmidt algorithm is proposed for the purpose. In addition, the radial basis function networks trained by the proposed training methods are compared with the X.D. He A Lapedes's model and the radial basis function network by nonrecursive training method. Through this comparison, an improved radial basis function network for predicting chaotic time series is presented. (author). 17 refs., 8 figs., 3 tabs.

• 한국경영정보학회:학술대회논문집
• /
• 2007.06a
• /
• pp.608-613
• /
• 2007

Prediction of corporate bankruptcies has long been an important topic and has been studied extensively in the finance and management literature because it is an essential basis for the risk management of financial institutions. Recently, support vector machines (SVMs) are becoming popular as a tool for bankruptcy prediction because they use a risk function consisting of the empirical error and a regularized term which is derived from the structural risk minimization principle. In addition, they don't require huge training samples and have little possibility of overfitting. However. in order to Use SVM, a user should determine several factors such as the parameters ofa kernel function, appropriate feature subset, and proper instance subset by heuristics, which hinders accurate prediction results when using SVM In this study, we propose a novel hybrid SVM classifier with simultaneous optimization of feature subsets, instance subsets, and kernel parameters. This study introduces genetic algorithms (GAs) to optimize the feature selection, instance selection, and kernel parameters simultaneously. Our study applies the proposed model to the real-world case for bankruptcy prediction. Experimental results show that the prediction accuracy of conventional SVM may be improved significantly by using our model.

• Journal of the Korean Society of Safety
• /
• v.31 no.4
• /
• pp.48-54
• /
• 2016

The prediction of various emissions from coal combustion is an important subject of researchers and engineers because of environmental consideration. Therefore, the development of the models for predicting pollutants very fast has received much attention from international research community, especially in the field of safety assessment. In this work, response surface method was introduced as a design of experiment, and the database for RSM was set with the numerical simulation of a drop tube furnace (DTF) to predict the spatial distribution of pollutant concentrations as well as final ones. The distribution of carbon dioxide in DTF was assumed to have Boltzman function, and the resulted function with parameters of a high $R^2$ value facilitates predicting an accurate distribution of $CO_2$. However, CO distribution had a difference near peak concentration when Gaussian function was introduced to simulate the CO distribution. It might be mainly due to the anti-symmetry of the CO concentration in DTF, and hence Extreme function was used to permit the asymmetry. The application of Extreme function enhanced the regression accuracy of parameters and the prediction was in a fairly good agreement with the new experiments. These results promise the wide use of statistical models for the quantitative safety assessment.

• Journal of Chest Surgery
• /
• v.28 no.4
• /
• pp.371-375
• /
• 1995

Surgical resection of lung cancer or other disease is recently required in patients with severely impaired lung function resulting from chronic obstructive pulmonary disease or disease extension. So prediction of pulmonary function after lung resection is very important in thoracic surgeon. We studied the accuracy of the prediction of postoperative pulmonary function using perfusion lung scan with 99m technetium macroaggregated albumin in 22 patients who received the pneumonectomy. The linear regression line derived from correlation between predicting[X and postoperative measured[Y values of FEV1 and FVC in patients are as follows: 1 Y[ml =0.713X + 381 in FEV1 [r=0.719 ,[P<0.01 2 Y[ml =0.645X + 556 in FVC [r=0.675 ,[P<0.01 In conclusion,the perfusion lung scan is noninvasive and very accurate for predicting postpneumonectomy pulmonary function.

• Journal of Applied Reliability
• /
• v.15 no.1
• /
• pp.44-51
• /
• 2015

In this study, we propose reliability prediction of a satellite by function analysis. To do so, the intended functions of the satellite are derived from using function structure block diagram, and defined as main, sub, and detailed functions. Furthermore, in order to generate function and reliability structure table, reliability model rule, duty cycle, and types of switch are assigned to the classified functions. This study also establishes reliability block diagram and mathematical reliability models to schematize the relationship among the functions. The reliability of the classified function is estimated by calculating the failure rate of parts comprising them. Finally, we apply the proposed method to a small satellite as a case study. The result shows that the reliability for the detailed function and the sub function as well as the main function could be predicted quantitatively and accurately by the proposed approach.

• Proceedings of the KIEE Conference
• /
• 2005.10b
• /
• pp.622-624
• /
• 2005

This paper proposes a time series prediction method for the nonllinear system using the fuzzy system and its genetic algorithm, At first, we obtain the optimal fuzzy membership function using the genetic algorithm. With the optimal fuzzy rules and its input differences, a better time prediction series system may be obtained. We obtain a good result for the time prediction of the electric load.

• Journal of Institute of Control, Robotics and Systems
• /
• v.17 no.11
• /
• pp.1173-1178
• /
• 2011

The power generated by wind turbines changes rapidly because of the continuous fluctuation of wind speed and direction. It is important for the power industry to have the capability to predict the changing wind power. In this paper, neural network based wind power prediction scheme which uses wind speed and direction is considered. In order to get a better prediction result, compression function which can be applied to the measurement data is introduced. Empirical data obtained from wind farm located in Kunsan is considered to verify the performance of the compression function.

• Composites Research
• /
• v.12 no.4
• /
• pp.1-7
• /
• 1999

Fatigue life prediction of matrix dominated composite laminates, which have a nonlinear stress/strain response, was studied analytically and experimentally. A stress function describing the relation of initial fatigue modulus and elastic modulus was used in order to consider the material nonlinearty. New modified fatigue life prediction equation was suggested based on the fatigue modulus and reference modulus concept as a function of applied stress. The prediction was verified by torsional fatigue test using crossply carbon/epoxy laminate tubes. It was shown that the proposed equation has wide applicability and good agreement with experimental data.

• International Journal of Contents
• /
• v.9 no.4
• /
• pp.11-15
• /
• 2013

"Protein Folding Problem" is considered to be one of the "Great Challenges of Computer Science" and prediction of disordered protein is an important part of the protein folding problem. Machine learning models can predict the disordered structure of protein based on its characteristic of "learning from examples". Among many machine learning models, we investigate the possibility of multilayer perceptron (MLP) as the predictor of protein disorder. The investigation includes a single hidden layer MLP, multi hidden layer MLP and the hierarchical structure of MLP. Also, the target node cost function which deals with imbalanced data is used as training criteria of MLPs. Based on the investigation results, we insist that MLP should have deep architectures for performance improvement of protein disorder prediction.