• Journal of Information Technology Services
• /
• v.21 no.4
• /
• pp.91-103
• /
• 2022

This study proposes a model for predicting contracted power using electric power data collected in real time from convenience stores nationwide. By optimizing the prediction model using machine learning, it will be possible to predict the contracted power required to renew the contract of the existing convenience store. Contracted power is predicted through the XGBoost regression model. For the learning of XGBoost model, the electric power data collected for 16 months through a real-time monitoring system for convenience stores nationwide were used. The hyperparameters of the XGBoost model were tuned using the GridesearchCV, and the main features of the prediction model were identified using the xgb.importance function. In addition, it was also confirmed whether the preprocessing method of missing values and outliers affects the prediction of reduced power. As a result of hyperparameter tuning, an optimal model with improved predictive performance was obtained. It was found that the features of power.2020.09, power.2021.02, area, and operating time had an effect on the prediction of contracted power. As a result of the analysis, it was found that the preprocessing policy of missing values and outliers did not affect the prediction result. The proposed XGBoost regression model showed high predictive performance for contract power. Even if the preprocessing method for missing values and outliers was changed, there was no significant difference in the prediction results through hyperparameters tuning.

• Journal of Communications and Networks
• /
• v.5 no.1
• /
• pp.65-72
• /
• 2003

Internet roundtrip delay/time (RTT) prediction plays an important role in detecting packet losses in reliable transport protocols for traditional web applications and determining proper transmission rates in many rate-based TCP-friendly protocols for Internet-based real-time applications. The widely adopted autoregressive and moving average (ARMA) model with fixed-parameters is shown to be insufficient for all scenarios due to its intrinsic limitation that it filters out all high-frequency components of RTT dynamics. In this paper, we introduce a novel parameter-varying RTT model for Internet roundtrip time prediction based on the information theory and the maximum entropy principle (MEP). Since the coefficients of the proposed RTT model are updated dynamically, the model is adaptive and it tracks RTT dynamics rapidly. The results of our experiments show that the MEP algorithm works better than the ARMA method in both RTT prediction and RTO estimation.

• Proceedings of the Korea Committee for Ocean Resources and Engineering Conference
• /
• 2000.10a
• /
• pp.101-108
• /
• 2000

In this paper, the real-time prediction of high temperature creep life was carried out for the friction welded joints of dissimilar heat resisting steels(SUH3-SUH35). Various life prediction methods such as LMP(Larson-Miller Parameter) and ISM(initial strain method) were applied : The creep behaviors of those steels and the welds under static load were examined by ISM combined with LMP at 500, 600 and $700^{\circ}C$, and the relationship between these two methods was investigated. A real-time creep life( $t_{r}$ , hr) prediction equation by initial strain($\varepsilon$$_{0}$ , %) under any creep stress ($\sigma$, MPa) at any high temperature(T, K) was developed as follows : $t_{r}$ =$\alpha$$\varepsilon$$_{0}$ $^{\beta}$$\sigma$$^{1}$ where, (equation omitted) for SUH3-SUH35 friction weld of =16mm and =20mm, respectively.

• Journal of Ocean Engineering and Technology
• /
• v.15 no.2
• /
• pp.46-52
• /
• 2001

In this paper, the real-time prediction of high temperature creep life was carried out for the friction welded joints of dissimilar heat resisting steels (SUH3-SUH35). various life prediction method such as LMP (Larson_miller Parameter) and ISM (initial strain method) were applied. The creep behaviors of those steels and the welds under static load were examined by ISM combined with LMP at 500, 600 and $700^{\circ}C$, and the relationship between these two methods was investigated. A real-time creep lie (tr, hr) prediction equation by initial strain (${\varepsilon}_0$, %) under any creep stress ($\sigma$, MPa) at any high temperature (T, K) was developed as follows: $t_r={\alpha}{\varepsilon}_0^{\beta}{\sigma}^{-1}$ where, ${\phi}=16: {\alpha}=10^{51.412-0.104T+5.375{\times}10^5T^2}$, $ {\beta}=-83.989+0.180T-9.957{\times}10^{-5}T^2,{\phi}=20:$ ${\alpha}=10^{69.910-0.146T+7.744{\times}10^{-5}T^2$, ${\beta}=-51.442+0.105T-5.595{\times}10^{-5}T^2$ for SUH3-SUH35 friction weld of =16mm and 20mm, respectively.

• Journal of Environmental Science International
• /
• v.6 no.3
• /
• pp.205-211
• /
• 1997

For the prediction of multi-site rainfall with radar data and ground meteorological data, a rainfall prediction model was proposed, which uses the neural network theory, a kind of artifical Intelligence technique. The Input layer of the prediction model was constructed with current ground meteorological data, their variation, moving vectors of rain- fall field and digital terrain of the measuring site, and the output layer was constructed with the predicted rainfall up to 3 hours. In the application of the prediction model to the Pyungchang river basin, the learning results of neural network prediction model showed more Improved results than the parameter estimation results of an existing physically based model. And the proposed model comparisonally well predicted the time distribution of ralnfall.

• KIPS Transactions on Computer and Communication Systems
• /
• v.11 no.2
• /
• pp.67-72
• /
• 2022

Recently, various intelligent application services using artificial intelligence are being actively developed. In particular, research on artificial intelligence-based real-time prediction services is being actively conducted in the manufacturing industry, and the demand for artificial intelligence services that can detect and predict fire and odors is very high. However, most of the existing detection and prediction systems do not predict the occurrence of fires and odors, but rather provide detection services after occurrence. This is because AI-based prediction service technology is not applied in existing systems. In addition, fire prediction, odor detection and odor level prediction services are services with ultra-low delay characteristics. Therefore, in order to provide ultra-low-latency prediction service, edge computing technology is combined with artificial intelligence models, so that faster inference results can be applied to the field faster than the cloud is being developed. Therefore, in this paper, we propose an LSTM algorithm-based learning model that can be used for fire prediction and odor detection/prediction, which are most required in the manufacturing industry. In addition, the proposed learning model is designed to be implemented in edge devices, and it is proposed to receive real-time sensor data from the IoT terminal and apply this data to the inference model to predict fire and odor conditions in real time. The proposed model evaluated the prediction accuracy of the learning model through three performance indicators, and the evaluation result showed an average performance of over 90%.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.19 no.7
• /
• pp.1568-1576
• /
• 2015

This paper proposes a prediction reputation system for the anti-SPIT solution in real-time VoIP environment. Increased accuracy of the determination as to whether spam or not by deriving a threshold based on SPIT presence in the existing paper. The existing schemes need to get the user's feedback and/or have experienced the time delay and overload as session initiates due to real-time operation. To solve these problems, the proposed scheme predicts the reputation through the statistical analysis based on the period of session initiation of each caller and the call duration of each receiver. As per the second mentioned problem, this scheme performs the prediction before session initiation, therefore, it's proper for real-time VoIP environment.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2017.05a
• /
• pp.194-194
• /
• 2017

Flooding is one of the most serious and frequently occurred natural disaster at many regions around the world. Especially, under the climate change impact, it is more and more increasingly trend. To reduce the flood damage, flood forecast and its accuracy analysis are required. This study is conducted to analyze the accuracy of the real-time flood forecasting of a coupled meteo-hydrological model for the Han River basin, South Korea. The LDAPS (Local Data Assimilation and Prediction System) products with the spatial resolution of 1.5km and lead time of 36 hours are extracted and used as inputs for the SURR (Sejong University Rainfall-Runoff) model. Three statistical criteria consisting of CC (Corelation Coefficient), RMSE (Root Mean Square Error) and ME (Model Efficiency) are used to evaluate the performance of this couple. The results are expected that the accuracy of the flood forecasting reduces following the increase of lead time corresponding to the accuracy reduction of LDAPS rainfall. Further study is planed to improve the accuracy of the real-time flood forecasting.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.05a
• /
• pp.547-549
• /
• 2021

In Artificial Intelligence analysis, the process of creating a model and verifying it is a task that requires computational processing time because it is Batch Processing performed with already generated data. We need to model, validate, and predict real-time data, such as stocks and defense information, with data generated directly in front of us. As a solution to this, we solve it by applying techniques to segment the data required for artificial intelligence modeling tasks in order of time processing and distribute the data across multiple processes.

• ETRI Journal
• /
• v.18 no.2
• /
• pp.61-74
• /
• 1996

We propose two new rate control schemes to improve MPEG-2 rate control in view of visual quality when scene changes happen. Two proposed schemes are characterized by real-time and non real-time improvement to reduce the impact of scene changes. We also propose a new target-bit prediction method using spatial activity of pictures and present a simple and efficient scene change detection scheme using signed difference of mean absolute difference (MAD). Computer simulation results show that the proposed real-time algorithm effectively alleviates visual quality degradation after scene changes. The proposed non real-time algorithm gives maximum 2 dB improvement in peak signal-to-noise ratio (PSNR) at a scene-changed picture, compared with MPEG-2 rate control scheme and it shows better quality than the real-time one.