• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.61 no.11
• /
• pp.1682-1688
• /
• 2012

This paper introduces a GP(Genetic Programming) based robust technique for temperature compensation in short-range prediction. Development of an efficient MOS(Model Output Statistics) is necessary to correct systematic errors of the model, because forecast models do not reliably determine weather conditions. Most of MOS use a linear regression to compensate a prediction model, therefore it is hard to manage an irregular nature of prediction. In order to solve the problem, a nonlinear and symbolic regression method using GP is suggested. The purpose of this study is to evaluate the accuracy of the estimation by a GP based nonlinear MOS for 3 days temperatures in Korean regions. This method is then compared to the UM model and has shown superior results. The training period of 2007-2009 summer is used, and the data of 2010 summer is adopted for verification.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.30 no.1
• /
• pp.55-74
• /
• 2005

Bankruptcy prediction has drawn a lot of research interests in previous literature, and recent studies have shown that machine learning techniques achieved better performance than traditional statistical ones. This paper employs a relatively new machine learning technique, support vector machines (SVMs). to bankruptcy prediction problem in an attempt to suggest a new model with better explanatory power and stability. To serve this purpose, we use grid search technique using 5-fold cross-validation to find out the optimal values of the parameters of kernel function of SVM. In addition, to evaluate the prediction accuracy of SVM. we compare its performance with multiple discriminant analysis (MDA), logistic regression analysis (Logit), and three-layer fully connected back-propagation neural networks (BPNs). The experiment results show that SVM outperforms the other methods.

• The Journal of Engineering Geology
• /
• v.20 no.3
• /
• pp.257-269
• /
• 2010

We analyzed patent trends regarding stabilization technology for steep slope hazards, focusing on patents applied for and registered in Korea, the USA, Japan, and Europe. The technology was classified into four groups at the second classification step: prediction techniques, instrumentation techniques, countermeasure/reinforcement/mitigation techniques, and laboratory tests. A total of 2,134 patents were selected for the final effective analysis. As a result of portfolio analysis using the correlation between the number of patents and the applicant for each patent, the Korean and USA situations were classified as belonging to the developing period, and the Japanese and European situations were classified as belonging to the ebbing period. In particular, patent activity in Korea has been enlivened by government-led research. As a result of technology analysis at the second classification step, prediction techniques arising from Japan are evaluated as a competitive power technique, and laboratory tests arising from the USA are evaluated as a competitive power technique. However, prediction techniques and laboratory tests arising from Korea are evaluated as a blank technique. According to the prediction results regarding future research and developments, a new finite element analysis method and a numerical model should be established as part of prediction techniques, as well as sensors, and hazard prediction should be developed by integrating information and equipment using IT technology as part of instrumentation techniques. In addition, improvements to existing structures for erosion control and the development of new slope-reinforcement methods are required as part of countermeasure/reinforcement/mitigation techniques, and new laboratory apparatus and methods with an optimizing structure should be developed as part of laboratory tests.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.31 no.8A
• /
• pp.745-751
• /
• 2006

The transmitter requires knowledge of the channel status information in order to adopt the adaptive modulation and coding scheme(AMC) for OFDM system. But in the outdoor environment which the users have high mobility, the channel status information from the users is outdated, so that it induces the degradation of system throughput and packet error rate(PER) performance. To solve this problem, researches about applying channel prediction technique to the AMC scheme have been proceeded. Most channel prediction techniques assume that there is no channel variation in the predefined time duration, e.g., a slot. As a result, those techniques cannot compensate the degradation of PER performance resulting from the rapid variation of channel during the slot duration. This paper introduces a novel channel prediction technique for OFDM/FDD system to support adaptive modulation and coding scheme over rapidly time-varying multipath fading channel. The proposed channel prediction technique considers the time-varying nature of channel during the slot duration. Simulation results show that the AMC scheme of OFDM/FDD system utilizing the proposed channel prediction technique can guarantee the target PER of 1% without any loss of system throughput compared with the case supported by the conventional channel prediction under ITU-R Veh A 30km/h.

• Proceedings of the KSRS Conference
• /
• v.2
• /
• pp.635-638
• /
• 2006

Regression of conventional prediction techniques in data mining uses the model which is generated from the training step. This model is applied to new input data without any change. If this model is applied directly to time series, the rate of prediction accuracy will be decreased. This paper proposes an incremental regression for time series prediction like typhoon track prediction. This technique considers the characteristic of time series which may be changed over time. It is composed of two steps. The first step executes a fractional process for applying input data to the regression model. The second step updates the model by using its information as new data. Additionally, the model is maintained by only recent data in a queue. This approach has the following two advantages. It maintains the minimum information of the model by using a matrix, so space complexity is reduced. Moreover, it prevents the increment of error rate by updating the model over time. Accuracy rate of the proposed method is measured by RME(Relative Mean Error) and RMSE(Root Mean Square Error). The results of typhoon track prediction experiment are performed by the proposed technique IMLR(Incremental Multiple Linear Regression) is more efficient than those of MLR(Multiple Linear Regression) and SVR(Support Vector Regression).

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.41 no.3
• /
• pp.154-161
• /
• 2018

This study conducted to predict the stock market prices based on the assumption that internet news articles might have an impact and effect on the rise and fall of stock market prices. The internet news articles were tested to evaluate the accuracy by comparing predicted values of the actual stock index and the forecasting models of the companies. This paper collected stock news from the internet, and analyzed and identified the relationship with the stock price index. Since the internet news contents consist mainly of unstructured texts, this study used text mining technique and multiple regression analysis technique to analyze news articles. A company H as a representative automobile manufacturing company was selected, and prediction models for the stock price index of company H was presented. Thus two prediction models for forecasting the upturn and decline of H stock index is derived and presented. Among the two prediction models, the error value of the prediction model (1) is low, and so the prediction performance of the model (1) is relatively better than that of the prediction model (2). As the further research, if the contents of this study are supplemented by real artificial intelligent investment decision system and applied to real investment, more practical research results will be able to be developed.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.56 no.12
• /
• pp.2251-2256
• /
• 2007

In this paper, the wavelet transform based lossless image compression algorithm is proposed. The proposed algorithm transforms the input image using 9/7 ICFB and S+P filter, and eliminates the spacious correlation of the subband coefficients, applying the context modeling predictive technique based on the multi-resolution structure and the feedback of the prediction error. The prediction context exploits the subordination and direction property of the different level subband in the vertical, horizontal, and diagonal subband coefficients. The simulation result of the high frequency images such as PEPPERS, BOAT, and AIRPLANE shows that the proposed algorithm efficiently predicts the edge area of each multi-resolution subband.

• KEPCO Journal on Electric Power and Energy
• /
• v.5 no.3
• /
• pp.157-163
• /
• 2019

K-BEMS System was introduced to reduce peak load and to save total energy of the 120 buildings that KEPCO headquarter and branch offices use. K-BEMS system is composed of PV, battery, and hybrid PCS. In this system, ESS, PV, lighting is used to save building energy based on demand prediction. Currently, neural network technique for short past data is applied to demand prediction, and fixed scheduling method by operator for ESS charging/discharging is used. To enhance this system, KEPCO research institute has carried out this K-BEMS research project for 3 years since January 2016. As the result of this project, we developed new real-time highly reliable building demand prediction technique with error free and optimized automatic ESS charging/discharging technique. Through several field test, we can certify the developed algorithm performance successfully. So we will describe the details in this paper.

• Geophysics and Geophysical Exploration
• /
• v.3 no.3
• /
• pp.83-87
• /
• 2000

The noise prediction filter using a local/remote reference was developed to obtain a high quality data from seismic surveys over the area where seismic transmission power is limited. The method used in the noise prediction filter is a 3-layer neural network whose algorithm is backpropagation. A NRF (Noise Reduction Factor) value of about 3.0 was obtained with appling training and test data to the trained noise prediction filter. However, the scaling technique generally used for minimizing EM noise from electric and electromagnetic data cannot reduce seismic noise, since the technique can allow only amplitude difference between two time series measured at the primary and reference sites.

• Journal of the Korean Society of Propulsion Engineers
• /
• v.5 no.2
• /
• pp.38-43
• /
• 2001

In this study, the technique of the performance prediction on the finocyl-type dual-thrust rocket motor is developed, and the predicted data are compared with those of the static firing tests. The prediction is carried out with the separate calculations of the grain burning area and the performance of the rocket motor. When predicting the performance of the dual-thrust rocket motor, the different correction factors should be used at the boosting and sustaining phases. Otherwise, an error of prediction will follow. Reprediction using the separate correction factors shows good agreement with the test data within 0.5% error.