• Journal of Korea Water Resources Association
• /
• v.55 no.8
• /
• pp.565-575
• /
• 2022

Recently, in the field of water resource engineering, interest in predicting time series water levels and flow rates using deep learning technology that has rapidly developed along with the Fourth Industrial Revolution is increasing. In addition, although water-level and flow-rate prediction have been performed using the Long Short-Term Memory (LSTM) model and Gated Recurrent Unit (GRU) model that can predict time-series data, the accuracy of flow-rate prediction in rivers with rapid temporal fluctuations was predicted to be very low compared to that of water-level prediction. In this study, the Paldang Bridge Station of the Han River, which has a large flow-rate fluctuation and little influence from tidal waves in the estuary, was selected. In addition, time-series data with large flow fluctuations were selected to collect water-level and flow-rate data for 2 years and 7 months, which are relatively short in data length, to be used as training and prediction data for the LSTM and GRU models. When learning time-series water levels with very high time fluctuation in two models, the predicted water-level results in both models secured appropriate accuracy compared to observation water levels, but when training rapidly temporal fluctuation flow rates directly in two models, the predicted flow rates deteriorated significantly. Therefore, in this study, in order to accurately predict the rapidly changing flow rate, the water-level data predicted by the two models could be used as input data for the rating curve to significantly improve the prediction accuracy of the flow rates. Finally, the results of this study are expected to be sufficiently used as the data of flood warning system in urban rivers where the observation length of hydrological data is not relatively long and the flow-rate changes rapidly.

• 한국경영정보학회:학술대회논문집
• /
• 2007.06a
• /
• pp.571-575
• /
• 2007

For better predictions and classifications in customer recommendation, this study proposes an integrative model that efficiently combines the currently-in-use statistical and artificial intelligence models. In particular, by integrating the models such as Association Rule, Connection Frequency Matrix, and Rule Induction, this study suggests an integrative prediction model.

• Journal of applied mathematics & informatics
• /
• v.6 no.2
• /
• pp.599-609
• /
• 1999

Neural networks are explored as an alternative to a regres-sion model for prediction of the number of daily household vehicular trips. This study focuses on contrasting a neural network model with a regression model in term of variable selection as well as the appli-cation of these models for prediction of extreme observations, The differences in the models regarding data transformation variable selec-tion and multicollinearity are considered. The results indicate that the neural network model is a viable alternative to the regression model for addressing both messy data problems and limitation in variable structure specification.

• Asian Pacific Journal of Cancer Prevention
• /
• v.13 no.3
• /
• pp.927-930
• /
• 2012

Background and Objectives: Artificial neural networks (ANNs) are flexible and nonlinear models which can be used by clinical oncologists in medical research as decision making tools. This study aimed to predict distant metastasis (DM) of colorectal cancer (CRC) patients using an ANN model. Methods: The data of this study were gathered from 1219 registered CRC patients at the Research Center for Gastroenterology and Liver Disease of Shahid Beheshti University of Medical Sciences, Tehran, Iran (January 2002 and October 2007). For prediction of DM in CRC patients, neural network (NN) and logistic regression (LR) models were used. Then, the concordance index (C index) and the area under receiver operating characteristic curve (AUROC) were used for comparison of neural network and logistic regression models. Data analysis was performed with R 2.14.1 software. Results: The C indices of ANN and LR models for colon cancer data were calculated to be 0.812 and 0.779, respectively. Based on testing dataset, the AUROC for ANN and LR models were 0.82 and 0.77, respectively. This means that the accuracy of ANN prediction was better than for LR prediction. Conclusion: The ANN model is a suitable method for predicting DM and in that case is suggested as a good classifier that usefulness to treatment goals.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.44 no.4
• /
• pp.227-233
• /
• 2021

Predictive maintenance has been one of important applications of data science technology that creates a predictive model by collecting numerous data related to management targeted equipment. It does not predict equipment failure with just one or two signs, but quantifies and models numerous symptoms and historical data of actual failure. Statistical methods were used a lot in the past as this predictive maintenance method, but recently, many machine learning-based methods have been proposed. Such proposed machine learning-based methods are preferable in that they show more accurate prediction performance. However, with the exception of some learning models such as decision tree-based models, it is very difficult to explicitly know the structure of learning models (Black-Box Model) and to explain to what extent certain attributes (features or variables) of the learning model affected the prediction results. To overcome this problem, a recently proposed study is an explainable artificial intelligence (AI). It is a methodology that makes it easy for users to understand and trust the results of machine learning-based learning models. In this paper, we propose an explainable AI method to further enhance the explanatory power of the existing learning model by targeting the previously proposedpredictive model [5] that learned data from a core facility (Hyper Compressor) of a domestic chemical plant that produces polyethylene. The ensemble prediction model, which is a black box model, wasconverted to a white box model using the Explainable AI. The proposed methodology explains the direction of control for the major features in the failure prediction results through the Explainable AI. Through this methodology, it is possible to flexibly replace the timing of maintenance of the machine and supply and demand of parts, and to improve the efficiency of the facility operation through proper pre-control.

• Structural Engineering and Mechanics
• /
• v.53 no.5
• /
• pp.921-938
• /
• 2015

Realistic prediction of concrete creep is of crucial importance for durability and long-term serviceability of concrete structures. To date, research about the behaviour of self-compacting concrete (SCC) members, especially concerning the long-term performance, is rather limited. SCC is quite different from conventional concrete (CC) in mixture proportions and applied materials, particularly in the presence of aggregate which is limited. Hence, the realistic prediction of creep strains in SCC is an important requirement for the design process of this type of concrete structures. This study reviews the accuracy of the conventional concrete (CC) creep prediction models proposed by the international codes of practice, including: CEB-FIP (1990), ACI 209R (1997), Eurocode 2 (2001), JSCE (2002), AASHTO (2004), AASHTO (2007), AS 3600 (2009). Also, SCC creep prediction models proposed by Poppe and De Schutter (2005), Larson (2007) and Cordoba (2007) are reviewed. Further, new creep prediction model based on the comprehensive analysis on both of the available models i.e. the CC and the SCC is proposed. The predicted creep strains are compared with the actual measured creep strains in 55 mixtures of SCC and 16 mixtures of CC.

• Earthquakes and Structures
• /
• v.8 no.5
• /
• pp.957-976
• /
• 2015

Recently, two energy-based response parameters, i.e., the absolute and the relative elastic input energy equivalent velocity, have been receiving a lot of research attention. Several studies, in fact, have demonstrated the potential of these intensity measures in the prediction of the seismic structural response. Although some ground motion prediction equations have been developed for these parameters, they only provide marginal distributions without information about the joint occurrence of the spectral values at different periods. In order to build new prediction models for the two equivalent velocities, a large set of ground motion records is used to calculate the correlation coefficients between the response spectral values corresponding to different periods and components of the ground motion. Then, functional forms adopted in models from the literature are calibrated to fit the obtained data. A new functional form is proposed to improve the predictions of the considered models from the literature. The components of the ground motion considered in this study are the two horizontal ones only. Potential uses of the proposed equations in addition to the prediction of the correlation coefficients of the equivalent velocity spectral values are shown, such as the prediction of derived intensity measures and the development of conditional mean spectra.

• YAKHAK HOEJI
• /
• v.59 no.1
• /
• pp.29-39
• /
• 2015

In recent years, physiologically based pharmacokinetic (PBPK) modeling has been widely used in pharmaceutical industries as well as regulatory health authorities for drug discovery and development. Several application areas of PBPK have been introduced so far including drug-drug interaction prediction, transporter-mediated interaction prediction, and pediatric PK prediction. The purpose of this review is to introduce PBPK and illustrates one of its application areas, particularly pediatric PK prediction by utilizing existing adult PK data and in vitro data. The evaluation of the initial PBPK for adult was done by comparing with experimental PK profiles and the scaling from adult to pediatric was conducted using age-related changes in size such as tissue compartments, and protein binding etc. Sotalol and lorazepam were selected in this review as model drugs for this purpose and were re-evaluated using the PBPK models by GastroPlus$^{(R)}$. The challenges and strategies of PBPK models using adult PK data as well as appropriate in vitro assay data for extrapolating pediatric PK at various ages were also discussed in this paper.

• 한국전산유체공학회:학술대회논문집
• /
• 2004.10a
• /
• pp.47-51
• /
• 2004

Accurate Prediction of a supersonic missile base drag continues to defy even well-rounded CFD codes. In an effort to address the accuracy and predictability of the base drags, the influence of grid system and competitive turbulence models on the base drag is analyzed. Characteristics of some turbulence models is reviewed through incompressible turbulent flow over a flat plate, and performance for the base drag prediction of several turbulence models such as Baldwin-Lomax(B-L), Spalart-Allmaras(S-A), $\kappa-\epsilon$, $\kappa-\omega$ model is assessed. When compressibility correction is injected into the S-A model, prediction accuracy of the base drag is enhanced. The NSWC wind tunnel test data are utilized for comparison of CFD and semi-empirical codes on the accuracy of base drag predictability: they are about equal, but CFD tends to perform better. It is also found that, as angle of attack of a missile with control (ins increases, even the best CFD analysis tool we have lacks the accuracy needed for the base drag prediction.

• Computers and Concrete
• /
• v.5 no.3
• /
• pp.261-277
• /
• 2008

This paper presents methodologies for remaining life prediction of plain concrete structural components considering tension softening effect. Non-linear fracture mechanics principles (NLFM) have been used for crack growth analysis and remaining life prediction. Various tension softening models such as linear, bi-linear, tri-linear, exponential and power curve have been presented with appropriate expressions. A methodology to account for tension softening effects in the computation of SIF and remaining life prediction of concrete structural components has been presented. The tension softening effects has been represented by using any one of the models mentioned above. Numerical studies have been conducted on three point bending concrete structural component under constant amplitude loading. Remaining life has been predicted for different loading cases and for various tension softening models. The predicted values have been compared with the corresponding experimental observations. It is observed that the predicted life using bi-linear model and power curve model is in close agreement with the experimental values. Parametric studies on remaining life prediction have also been conducted by using modified bilinear model. A suitable value for constant of modified bilinear model is suggested based on parametric studies.