• Journal of Korea Multimedia Society
• /
• v.23 no.6
• /
• pp.747-755
• /
• 2020

In this paper, we deal with outlier data problems that occur when constructing a PM_2.5 fine dust forecasting system using a neural network. In general, when learning a neural network, some of the data are not helpful for learning, but rather disturbing. Those are called outlier data. When they are included in the training data, various problems such as overfitting occur. In building a PM_2.5 fine dust concentration forecasting system using neural network, we have found several outlier data in the training data. We, therefore, remove them, and then make learning 3 ways. Over_outlier model removes outlier data that target concentration is low, but the model forecast is high. Under_outlier model removes outliers data that target concentration is high, but the model forecast is low. All_outlier model removes both Over_outlier and Under_outlier data. We compare 3 models with a conventional outlier removal model and non-removal model. Our outlier removal model shows better performance than the others.

• Journal of Korea Multimedia Society
• /
• v.24 no.10
• /
• pp.1414-1424
• /
• 2021

In this paper, the importance of input factors of a DNN (Deep Neural Network) PM_2.5 forecasting model using LRP(Layer-wise Relevance Propagation) is analyzed, and forecasting performance is improved. Input factor importance analysis is performed by dividing the learning data into time and PM_2.5 concentration. As a result, in the low concentration patterns, the importance of weather factors such as temperature, atmospheric pressure, and solar radiation is high, and in the high concentration patterns, the importance of air quality factors such as PM_2.5, CO, and NO₂ is high. As a result of analysis by time, the importance of the measurement factors is high in the case of the forecast for the day, and the importance of the forecast factors increases in the forecast for tomorrow and the day after tomorrow. In addition, date, temperature, humidity, and atmospheric pressure all show high importance regardless of time and concentration. Based on the importance of these factors, the LRP_DNN prediction model is developed. As a result, the ACC(accuracy) and POD(probability of detection) are improved by up to 5%, and the FAR(false alarm rate) is improved by up to 9% compared to the previous DNN model.

• Journal of the Korean Society of Urban Environment
• /
• v.18 no.4
• /
• pp.531-546
• /
• 2018

Air quality forecasting system with Asian dust emissions was developed in East Asia, and $PM_{10}$ forecasting performance of chemical transport model with Asian dust emissions was validated and evaluated. The chemical transport model (CTM) with Asian dust emission was found to supplement $PM_{10}$ concentrations that had been under-estimated in China regions and improved statistics for performance of CTM, although the model were overestimated during some periods in China. In Korea, the prediction model adequately simulated inflow of Asian dust events on February 22~24 and March 16~17, but the model is found to be overestimated during no Asian dust event periods on April. However, the model supplemented $PM_{10}$ concentrations, which was underestimated in most regions in Korea and the statistics for performance of the models were improved. The $PM_{10}$ forecasting performance of air quality forecasting model with Asian dust emissions tends to improve POD (Probability of Detection) compared to basic model without Asian dust emissions, but A (Accuracy) has shown similar or decreased, and FAR (False Alarms) have increased during 2017.Therefore, the developed air quality forecasting model with Asian dust emission was not proposed as a representative $PM_{10}$ forecast model in South Korea.

• 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.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.43 no.4
• /
• pp.59-66
• /
• 2020

In this study, we consider the problem of forecasting the number of inbound foreigners visiting Korea. Forecasting tourism demand is an essential decision to plan related facilities and staffs, thus many studies have been carried out, mainly focusing on the number of inbound or outbound tourists. In order to forecast tourism demand, we use a seasonal ARIMA (SARIMA) model, as well as a SARIMAX model which additionally comprises an exogenous variable affecting the dependent variable, i.e., tourism demand. For constructing the forecasting model, we use a search procedure that can be used to determine the values of the orders of the SARIMA and SARIMAX. For the exogenous variable, we introduce factors that could cause the tourism demand reduction, such as the 9/11 attack, the SARS and MERS epidemic, and the deployment of THAAD. In this study, we propose a procedure, called Measuring Impact on Demand (MID), where the impact of each factor on tourism demand is measured and the value of the exogenous variable corresponding to the factor is determined based on the measurement. To show the performance of the proposed forecasting method, an empirical analysis was conducted where the monthly number of foreign visitors in 2019 were forecasted. It was shown that the proposed method can find more accurate forecasts than other benchmarks in terms of the mean absolute percentage error (MAPE).

• The Korean Journal of Applied Statistics
• /
• v.35 no.5
• /
• pp.645-655
• /
• 2022

In this paper, we investigate automatic functions for time series forecasting in R system and compare their performances. For the exponential smoothing models and ARIMA (autoregressive integrated moving average) models, we focus on the representative time series forecasting functions in R: forecast::ets(), forecast::auto.arima(), smooth::es() and smooth::auto.ssarima(). In order to compare their forecast performances, we use M3-Competiti on data consisting of 3,003 time series and adopt 3 accuracy measures. It is confirmed that each of the four automatic forecasting functions has strengths and weaknesses in the flexibility and convenience for time series modeling, forecasting accuracy, and execution time.

• The Journal of Industrial Distribution & Business
• /
• v.9 no.10
• /
• pp.83-89
• /
• 2018

Purpose - The purpose of this study examines when the optimism impact on financial asset price forecasting and the boundary condition of optimism in the financial asset price forecasting. People generally tend to optimistically forecast their future. Optimism is a nature of human beings and optimistic forecasting observed in daily life. But is it always observed in financial asset price forecasting? In this study, two factors were focused on considering whether the optimism that people have applied to predicting future performance of financial investment products (e.g., mutual fund). First, this study examined whether the degree of optimism varied depending on the direction of the prior price trend. Second, this study examined whether the degree of optimism varied according to the forecast period by dividing the future forecasted by people into three time horizon based on forecast period. Research design, data, and methodology - 2 (prior price trend: rising-up trend vs falling-down trend) × 3 (forecast time horizon: short term vs medium term vs long term) experimental design was used. Prior price trend was used between subject and forecast time horizon was used within subject design. 169 undergraduate students participated in the experiment. χ² analysis was used. In this study, prior price trend divided into two types: rising-up trend versus falling-down trend. Forecast time horizon divided into three types: short term (after one month), medium term (after one year), and long term (after five years). Results - Optimistic price forecasting and boundary condition was found. Participants who were exposed to falling-down trend did not make optimistic predictions in the short term, but over time they tended to be more optimistic about the future in the medium term and long term. However, participants who were exposed to rising-up trend were over-optimistic in the short term, but over time, less optimistic in the medium and long term. Optimistic price forecasting was found when participants forecasted in the long term. Exposure to prior price trends (rising-up trend vs falling-down trend) was a boundary condition of optimistic price forecasting. Conclusions - The results indicated that individuals were more likely to be impacted by prior price tends in the short term time horizon, while being optimistic in the long term time horizon.

• The Korean Journal of Applied Statistics
• /
• v.35 no.3
• /
• pp.385-393
• /
• 2022

In this paper, we consider univariate time series models that are well known in the field of forecasting and we study on forecasting performance for their simple combinations. The univariate time series models include exponential smoothing methods and ARIMA (autoregressive integrated moving average) models, their extended models, and non-seasonal and seasonal random walk models, which is frequently used as benchmark models for forecasting. The median and mean are simply used for the combination method, and the data set used for performance evaluation is M3-competition data composed of 3,003 various time series data. As results of evaluating the performance by sMAPE (symmetric mean absolute percentage error) and MASE (mean absolute scaled error), we assure that the simple combinations of the univariate models perform very well in the M3-competition dataset.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.32 no.1B
• /
• pp.9-20
• /
• 2012

Recently to overcome limitations of conceptual, hydrological and physics based models for flood stage forecasting, multiple linear regression model as one of data-driven models have been widely adopted for forecasting flood streamflow(stage). The objectives of this study are to compare performance of different multiple linear regression models according to regression coefficient estimation methods and determine most effective multiple linear regression flood stage forecasting models. To do this, the time scale was determined through the autocorrelation analysis of input data and different flood stage forecasting models developed using regression coefficient estimation methods such as LS(least square), WLS(weighted least square), SPW(stepwise) was applied to flood events in Jungrang stream. To evaluate performance of established models, fours statistical indices were used, namely; Root mean square error(RMSE), Nash Sutcliffe efficiency coefficient (NSEC), mean absolute error (MAE), adjusted coefficient of determination($R^{*2}$). The results show that the flood stage forecasting model using SPW(stepwise) parameter estimation can carry out the river flood stage prediction better in comparison with others, and the flood stage forecasting model using LS(least square) parameter estimation is also found to be slightly better than the flood stage forecasting model using WLS(weighted least square) parameter estimation.

• East Asian Journal of Business Economics (EAJBE)
• /
• v.1 no.1
• /
• pp.17-21
• /
• 2013

Purpose: Forecasting of petroleum consumption is useful in planning and management of petroleum production and control of air pollution. Research Design, Data and Methodology: ARMA models, sometimes called Box-Jenkins models after the iterative Box-Jenkins methodology usually used to estimate them, are typically applied to auto correlated time series data. Results: Petroleum consumption modeling plays a role key in big urban air pollution planning and management. In this study three models as, MLFF, MLFF with GARCH (1,1) and ARMA(1,1), have been investigated to model the petroleum consumption forecasts. Certain standard statistical parameters were used to evaluate the performance of the models developed in this study. Based upon the results obtained in this study and the consequent comparative analysis, it has been found that the MLFF with GARCH (1,1) have better forecasting results.. Conclusions: Survey of data reveals that deposit of government policies in recent yeas, petroleum consumption rises in Tehran and unfortunately more petroleum use causes to air pollution and bad environmental problems.