• The Transactions of the Korean Institute of Electrical Engineers P
• /
• v.67 no.4
• /
• pp.183-190
• /
• 2018

Predicting accurate electricity prices is an important task in the electricity trading market. To address the electricity price forecasting problem, various approaches have been proposed so far and it is known that linear regression-based approaches are the best. However, the use of such linear regression-based methods is limited due to low accuracy and performance. In traditional linear regression methods, it is not practical to find a nonlinear regression model that explains the training data well. If the training data is complex (i.e., small-sized individual data and large-sized features), it is difficult to find the polynomial function with n terms as the model that fits to the training data. On the other hand, as a linear regression model approximating a nonlinear regression model is used, the accuracy of the model drops considerably because it does not accurately reflect the characteristics of the training data. To cope with this problem, we propose a new electricity price forecasting method that divides the entire dataset to multiple split datasets and find the best linear regression models, each of which is the optimal model in each dataset. Meanwhile, to improve the performance of the proposed method, we modify the proposed localized linear regression method in the map and reduce way that is a framework for parallel processing data stored in a Hadoop distributed file system. Our experimental results show that the proposed model outperforms the existing linear regression model. Specifically, the accuracy of the proposed method is improved by 45% and the performance is faster 5 times than the existing linear regression-based model.

• Journal of Climate Change Research
• /
• v.7 no.2
• /
• pp.193-203
• /
• 2016

The importance of forecasting agricultural production is receiving attention while climate change is accelerating. This study suggested three types of crop yield forecasting model for major vegetable crops by using downscaled meteorological information of main production area on farmland level, which identified as limitation from previous studies. First, this study conducted correlation analysis with seven types of farm level downscaled meteorological informations and reported crop yield of main production area. After, we selected three types of meteorological factors which showed the highest relation with each crop species and regions. Parameters were deducted from meterological factor with high correlation but crop species number was neglected. After, crop yield of each crops was estimated by using the three suggested types of models. Chinese cabbage showed high accuracy in overall, while the accuracy of daikon and onion was quiet revised by neglecting the outlier. Chili and garlic showed differences by region, but Kyungbuk chili and Chungnam, Kyungsang garlic appeared significant accuracy. We also selected key meteorological factor of each crops which has the highest relation with crop yield. If the factor had significant relation with the quantity, it explains better about the variations of key meteorological factor. This study will contribute to establishing the methodology of future studies by estimating the crop yield of different species by using farmland meterological information and relatively simplify multiple linear regression models.

• Journal of Technology Innovation
• /
• v.17 no.1
• /
• pp.97-117
• /
• 2009

The Delphi is a popular technique for forecasting based on the opinions of experts. It is important to know how valid and reliable the technique is. In this paper, we analyze accuracy and precision of the Delphi in IT and BT of Korea, and also discuss the relationship between them. As a result of the analysis, the accuracy and precision of the forecasts partly have significant differences according to their area and degree of expertise. Besides, significant correlation between the accuracy and precision of forecasts with high expertise is found. The result indicates that the precision of forecasts can be a criterion of the accuracy of them.

• Journal of Distribution Science
• /
• v.12 no.2
• /
• pp.59-71
• /
• 2014

Purpose - In recent years, many firms have attempted various approaches to cope with the continual increase of aviation transportation. The previous research into freight charge forecasting models has focused on regression analyses using a few influence factors to calculate the future price. However, these approaches have limitations that make them difficult to apply into practice: They cannot respond promptly to small price changes and their predictive power is relatively low. Therefore, the current study proposes a freight charge-forecasting model using time series data instead a regression approach. The main purposes of this study can thus be summarized as follows. First, a proper model for freight charge using the autoregressive integrated moving average (ARIMA) model, which is mainly used for time series forecast, is presented. Second, a modified ARIMA model for freight charge prediction and the standard process of determining freight charge based on the model is presented. Third, a straightforward freight charge prediction model for practitioners to apply and utilize is presented. Research design, data, and methodology - To develop a new freight charge model, this study proposes the ARIMAC(p,q) model, which applies time difference constantly to address the correlation coefficient (autocorrelation function and partial autocorrelation function) problem as it appears in the ARIMA(p,q) model and materialize an error-adjusted ARIMAC(p,q). Cargo Account Settlement Systems (CASS) data from the International Air Transport Association (IATA) are used to predict the air freight charge. In the modeling, freight charge data for 72 months (from January 2006 to December 2011) are used for the training set, and a prediction interval of 23 months (from January 2012 to November 2013) is used for the validation set. The freight charge from November 2012 to November 2013 is predicted for three routes - Los Angeles, Miami, and Vienna - and the accuracy of the prediction interval is analyzed using mean absolute percentage error (MAPE). Results - The result of the proposed model shows better accuracy of prediction because the MAPE of the error-adjusted ARIMAC model is 10% and the MAPE of ARIMAC is 11.2% for the L.A. route. For the Miami route, the proposed model also shows slightly better accuracy in that the MAPE of the error-adjusted ARIMAC model is 3.5%, while that of ARIMAC is 3.7%. However, for the Vienna route, the accuracy of ARIMAC is better because the MAPE of ARIMAC is 14.5% and the MAPE of the error-adjusted ARIMAC model is 15.7%. Conclusions - The accuracy of the error-adjusted ARIMAC model appears better when a route's freight charge variance is large, and the accuracy of ARIMA is better when the freight charge variance is small or has a trend of ascent or descent. From the results, it can be concluded that the ARIMAC model, which uses moving averages, has less predictive power for small price changes, while the error-adjusted ARIMAC model, which uses error correction, has the advantage of being able to respond to price changes quickly.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2011.05a
• /
• pp.15-15
• /
• 2011

The radar observation system in Japan is operated by two governmental groups: Japan Meteorological Agency (JMA) and the Ministry of Land, Infrastructure, Transport and Tourism (MLIT) of Japan. The JMA radar observation network is comprised of 20 C-band radars (with a wavelength of 5.6 cm), which cover most of the Japan Islands and observe rainfall intensity and distribution. And the MLIT's radar observation system is composed of 26 C-band radars throughout Japan. The observed radar echo from each radar unit is first modified, and then sent to the National Bureau of Synthesis Process within the MLIT. Through several steps for homogenizing observation accuracy, including distance and elevation correction, synthesized rainfall intensity maps for the entire nation of Japan are generated every 5 minutes. The MLIT has recently launched a new radar observation network system designed for flash flood observation and forecasting in small river basins within urban areas. It is called the X-band multi parameter radar network, and is distinguished by its dual polarimetric wave pulses of short length (3cm). Attenuation problems resulting from the short wave length of radar echo are strengthened by polarimetric wavelengths and very dense radar networks. Currently, the network is established within four areas. Each area is observed using 3-4 X-band radars with very fine resolution in spatial (250 m) and temporal (1 minute intervals). This study provides a series of utilization procedures for the new input data into a real-time forecasting system. First of all, the accuracy of the X-band radar observation was determined by comparing its results with the rainfall intensities as observed by ground gauge stations. It was also compared with conventional C-band radar observation. The rainfall information from the new radar network was then provided to a distributed hydrologic model to simulate river discharges. The simulated river discharges were evaluated again using the observed river discharge to estimate the applicability of the new observation network in the context of operations regarding flood forecasting. It was able to determine that the newly equipped X-band polarimetric radar network shows somewhat improved observation accuracy compared to conventional C-band radar observation. However, it has a tendency to underestimate the rainfall, and the accuracy is not always superior to that of the C-band radar. The accuracy evaluation of the X-band radar observation in this study was conducted using only limited rainfall events, and more cases should be examined for developing a broader understanding of the general behavior of the X-band radar and for improving observation accuracy.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.27 no.6
• /
• pp.17-22
• /
• 2013

Fuzzy linear regression method has been used for short-term load forecasting of the special day in the previous researches. However, considerable load forecasting errors would be occurring if a special day is located on Saturday or Monday. In this paper, a new load forecasting method for the consecutive holidays is proposed with the consideration of the power demand variation rate. In the proposed method, a exponential smoothing model reflecting temperature is used to short-term load forecasting for Sunday during the consecutive holidays and then the loads of the special day during the consecutive holidays is calculated using the hourly power demand variation rate between the previous similar consecutive holidays. The proposed method is tested with 10 cases of the consecutive holidays from 2009 to 2012. Test results show that the average accuracy of the proposed method is improved about 2.96% by comparison with the fuzzy linear regression method.

• Journal of Intelligence and Information Systems
• /
• v.15 no.4
• /
• pp.37-51
• /
• 2009

One of the critical issues in the management of manufacturing companies is the efficient process of planning and operating service resources such as human, parts, and facilities, and it begins with the accurate service demand forecasting. In this research, service and sales data from the LCD monitor manufacturer is considered for an empirical study on Product Life Cycle (PLC) based service demand forecasting. The proposed PLC forecasting approach consists of four steps : understanding the basic statistics of data, clustering models using a self-organizing map, developing respective forecasting models for each segment, comparing the accuracy performance. Empirical experiments show that the PLC approach outperformed the traditional approaches in terms of root mean square error and mean absolute percentage error.

• Proceedings of the KSR Conference
• /
• 2011.10a
• /
• pp.1715-1721
• /
• 2011

For an efficient railroad operations the demand forecasting is required. Time series models can quickly forecast the future demand with fewer data. As well as the accuracy of forecasting is excellent compared to other methods. In this study is proposed the intervention ARIMA model for forecasting methods of KTX passenger demand. The intervention ARIMA model may reflect the intervention such as the Kyongbu high-speed rail project second phase. The simple seasonal ARIMA model is predicted to overestimate the KTX passenger demand. However, intervention ARIMA model is predicted the reasonable results. The KTX passenger demands were predicted to be a week units separated by the weekday and weekend.

• Asian Journal of Innovation and Policy
• /
• v.4 no.1
• /
• pp.35-54
• /
• 2015

The purpose of this study is to compare the valuation of technology firms in the KOSPI and KOSDAQ. This study analyzed 224 market reports for KOSDAQ firms and 602 reports for KOSPI firms. We compare the two markets under 3 definitions on the accuracy of stock price forecasting. Findings are as follows: Although PER multiples is the most used method of valuation, KOSDAQ valuation more heavily relies on the method than KOSPI valuation. In stock market, the period of earnings forecasting is mostly 2-3 years. Multiples of KOSDAQ is generally higher than those of KOSPI. Even for technology firms, valuation in KOSPI mostly relies on earnings of the company, but that in KOSDAQ mostly relies on relative price. In stock price forecasting, generally overestimation prevails. Moreover, forecasting of KOSPI reports is more accurate than that of KOSDAQ reports. ROE and COE of KOSDAQ firms are generally higher than those of KOSPI firms.

• The Korean Journal of Applied Statistics
• /
• v.25 no.2
• /
• pp.251-259
• /
• 2012

This paper applies forecasting models such as ARIMA, Holt-Winters and AR-GARCH models to analyze daily tourism data in Korea. To evaluate the performance of the models, we need single and double seasonal models that compare the RMSE and SE for a better accuracy of the forecasting models based on Armstrong (2001).