• Journal of Intelligence and Information Systems
• /
• v.19 no.1
• /
• pp.1-17
• /
• 2013

In this article, several types of hybrid forecasting models are suggested. In particular, hybrid models using the generalized additive model (GAM) are newly suggested as an alternative to those using neural networks (NN). The prediction performances of various hybrid and non-hybrid models are evaluated using simulated time series data. Five different types of seasonal time series data related to an additive or multiplicative trend are generated over different levels of noise, and applied to the forecasting evaluation. For the simulated data with only seasonality, the autoregressive (AR) model and the hybrid AR-AR model performed equivalently very well. On the other hand, if the time series data employed a trend, the SARIMA model and some hybrid SARIMA models equivalently outperformed the others. In the comparison of GAMs and NNs, regarding the seasonal additive trend data, the SARIMA-GAM evenly performed well across the full range of noise variation, whereas the SARIMA-NN showed good performance only when the noise level was trivial.

• Journal of Korea Water Resources Association
• /
• v.53 no.2
• /
• pp.97-105
• /
• 2020

Recently, Artificial Neural Network receives attention as a data prediction method. Among these, a Long Shot-term Memory (LSTM) model specialized for time-series data prediction was utilized as a prediction method of hydrological time series data. In this study, the LSTM model was constructed utilizing deep running open source library TensorFlow which provided by Google, to predict inflows of multipurpose dams. We predicted the inflow of the Yongdam Multipurpose Dam which is located in the upper stream of the Geumgang. The hourly flow data of Yongdam Dam from 2006 to 2018 provided by WAMIS was used as the analysis data. Predictive analysis was performed under various of variable condition in order to compare and analyze the prediction accuracy according to four learning parameters of the LSTM model. Root mean square error (RMSE), Mean absolute error (MAE) and Volume error (VE) were calculated and evaluated its accuracy through comparing the predicted and observed inflows. We found that all the models had lower accuracy at high inflow rate and hourly precipitation data (2006~2018) of Yongdam Dam utilized as additional input variables to solve this problem. When the data of rainfall and inflow were utilized together, it was found that the accuracy of the prediction for the high flow rate is improved.

• Journal of the Korean Society of Visualization
• /
• v.1 no.2
• /
• pp.58-70
• /
• 2003

Meteorological data contains observation and numerical weather prediction model output data. The computerized analysis and visualization of meteorological data often requires very high computing capability due to the large size and complex structure of the data. Because the meteorological data is frequently formed in multi-variables, 3-dimensional and time-series form, it is very important to visualize and analyze the data in 3D spatial domain in order to get more understanding about the meteorological phenomena. In this research, we developed interactive 3-dimensional visualization techniques for visualizing meteorological data on a PC environment such as volume rendering, iso-surface rendering or stream line. The visualization techniques developed in this research are expected to be effectively used as basic technologies not only for deeper understanding and more exact prediction about meteorological environments but also for scientific and spatial data visualization research in any field from which three dimensional data comes out such as oceanography, earth science, and aeronautical engineering.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.27 no.1
• /
• pp.48-55
• /
• 2022

This paper proposes a battery remaining useful life (RUL) prediction method using a deep learning-based EMD-CNN-LSTM hybrid method. The proposed method pre-processes capacity data by applying empirical mode decomposition (EMD) and predicts the remaining useful life using CNN-LSTM. CNN-LSTM is a hybrid method that combines convolution neural network (CNN), which analyzes spatial features, and long short term memory (LSTM), which is a deep learning technique that processes time series data analysis. The performance of the proposed remaining useful life prediction method is verified using the battery aging experiment data provided by the NASA Ames Prognostics Center of Excellence and shows higher accuracy than does the conventional method.

• Korean Journal of Construction Engineering and Management
• /
• v.22 no.3
• /
• pp.21-30
• /
• 2021

This study proposes a 5-month rebar price prediction method using the recursive prediction method of deep learning. This approach predicts a long-term point in time by repeating the process of predicting all the characteristics of the input data and adding them to the original data and predicting the next point in time. The predicted average accuracy of the rebar prices for one to five months is approximately 97.24% in the manner presented in this study. Through the proposed method, it is expected that more accurate cost planning will be possible than the existing method by supplementing the systematicity of the price estimation method through human experience and judgment. In addition, it is expected that the method presented in this study can be utilized in studies that predict long-term prices using time series data including building materials other than rebar.

• Information Systems Review
• /
• v.23 no.1
• /
• pp.23-43
• /
• 2021

Recently, many studies have been conducted to increase the accuracy of stock price prediction by analyzing candlestick charts using artificial intelligence techniques. However, these studies failed to consider the time-series characteristics of candlestick charts and to take into account the emotional state of market participants in data learning for stock price prediction. In order to overcome these limitations, this study produced input data by combining volatility index and candlestick charts to consider the emotional state of market participants, and used the data as input for a new method proposed on the basis of combining variantion autoencoder (VAE) and attention mechanisms for considering the time-series characteristics of candlestick chart. Fifty firms were randomly selected from the S&P 500 index and their stock prices were predicted to evaluate the performance of the method compared with existing ones such as convolutional neural network (CNN) or long-short term memory (LSTM). The results indicated the method proposed in this study showed superior performance compared to the existing ones. This study implied that the accuracy of stock price prediction could be improved by considering the emotional state of market participants and the time-series characteristics of the candlestick chart.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.10 no.4
• /
• pp.315-323
• /
• 2000

In this paper, we propose a method of constructing neural networks using bio-inpired emergent and evolutionary concepts. This method is algorithm that is based on the characteristics of the biological DNA and growth of plants, Here is, we propose a constructing method to make a DNA coding method for production rule of L-system. L-system is based on so-called the parallel rewriting nechanism. The DNA coding method has no limitation in expressing the produlation the rule of L-system. Evolutionary algotithms motivated by Darwinaian natural selection are population based searching methods and the high performance of which is highly dependent on the representation of solution space. In order to verify the effectiveness of our scheme, we apply it one step ahead prediction of Mackey-Glass time series, Sunspot data and KOSPI data.

• Korean Journal of Remote Sensing
• /
• v.35 no.6_3
• /
• pp.1209-1219
• /
• 2019

For consistent vegetation monitoring, it is necessary to generate time-series vegetation index datasets at fine temporal and spatial scales by fusing the complementary characteristics between temporal and spatial scales of multiple satellite data. In this study, we quantitatively and qualitatively analyzed the prediction accuracy of time-series change information extracted from spatio-temporal fusion models of multiple satellite data for vegetation monitoring. As for the spatio-temporal fusion models, we applied two models that have been widely employed to vegetation monitoring, including a Spatial and Temporal Adaptive Reflectance Fusion Model (STARFM) and an Enhanced Spatial and Temporal Adaptive Reflectance Fusion Model (ESTARFM). To quantitatively evaluate the prediction accuracy, we first generated simulated data sets from MODIS data with fine temporal scales and then used them as inputs for the spatio-temporal fusion models. We observed from the comparative experiment that ESTARFM showed better prediction performance than STARFM, but the prediction performance for the two models became degraded as the difference between the prediction date and the simultaneous acquisition date of the input data increased. This result indicates that multiple data acquired close to the prediction date should be used to improve the prediction accuracy. When considering the limited availability of optical images, it is necessary to develop an advanced spatio-temporal model that can reflect the suggestions of this study for vegetation monitoring.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2021.06a
• /
• pp.131-131
• /
• 2021

Deep learning models, especially those based on long short-term memory (LSTM), have presented their superiority in addressing time series data issues recently. This study aims to comprehensively evaluate the performance of deep learning models that belong to the supervised learning category in streamflow prediction. Therefore, six deep learning models-standard LSTM, standard gated recurrent unit (GRU), stacked LSTM, bidirectional LSTM (BiLSTM), feed-forward neural network (FFNN), and convolutional neural network (CNN) models-were of interest in this study. The Red River system, one of the largest river basins in Vietnam, was adopted as a case study. In addition, deep learning models were designed to forecast flowrate for one- and two-day ahead at Son Tay hydrological station on the Red River using a series of observed flowrate data at seven hydrological stations on three major river branches of the Red River system-Thao River, Da River, and Lo River-as the input data for training, validation, and testing. The comparison results have indicated that the four LSTM-based models exhibit significantly better performance and maintain stability than the FFNN and CNN models. Moreover, LSTM-based models may reach impressive predictions even in the presence of upstream reservoirs and dams. In the case of the stacked LSTM and BiLSTM models, the complexity of these models is not accompanied by performance improvement because their respective performance is not higher than the two standard models (LSTM and GRU). As a result, we realized that in the context of hydrological forecasting problems, simple architectural models such as LSTM and GRU (with one hidden layer) are sufficient to produce highly reliable forecasts while minimizing computation time because of the sequential data nature.

• KIPS Transactions on Software and Data Engineering
• /
• v.9 no.4
• /
• pp.129-136
• /
• 2020

PM2.5 which is a very tiny air particulate matter even smaller than PM10 has been issued in the environmental problem. Since PM2.5 can cause eye diseases or respiratory problems and infiltrate even deep blood vessels in the brain, it is important to predict PM2.5. However, it is difficult to predict PM2.5 because there is no clear explanation yet regarding the creation and the movement of PM2.5. Thus, prediction methods which not only predict PM2.5 accurately but also have the interpretability of the result are needed. To predict hourly PM2.5 of Seoul city, we propose a method using random forest with the adjusted bootstrap number from the time series ground data preprocessed on different sources. With this method, the prediction model can be trained uniformly on hourly information and the result has the interpretability. To evaluate the prediction performance, we conducted comparative experiments. As a result, the performance of the proposed method was superior against other models in all labels. Also, the proposed method showed the importance of the variables regarding the creation of PM2.5 and the effect of China.