• Journal of Korea Water Resources Association
• /
• v.57 no.9
• /
• pp.615-626
• /
• 2024

To accurately and efficiently monitor soil moisture (SM) across South Korea, this study developed a SM estimation model that integrates the cloud computing platform Google Earth Engine (GEE) and Automated Machine Learning (AutoML). Various spatial information was utilized based on Terra MODIS (Moderate Resolution Imaging Spectroradiometer) and the global precipitation observation satellite GPM (Global Precipitation Measurement) to test optimal input data combinations. The results indicated that GPM-based accumulated dry-days, 5-day antecedent average precipitation, NDVI (Normalized Difference Vegetation Index), the sum of LST (Land Surface Temperature) acquired during nighttime and daytime, soil properties (sand and clay content, bulk density), terrain data (elevation and slope), and seasonal classification had high feature importance. After setting the objective function (Determination of coefficient, R² ; Root Mean Square Error, RMSE; Mean Absolute Percent Error, MAPE) using AutoML for the combination of the aforementioned data, a comparative evaluation of machine learning techniques was conducted. The results revealed that tree-based models exhibited high performance, with Random Forest demonstrating the best performance (R² : 0.72, RMSE: 2.70 vol%, MAPE: 0.14).

• Geomechanics and Engineering
• /
• v.30 no.6
• /
• pp.551-564
• /
• 2022

Fragmenting the rock mass is considered as the most important work in open-pit mines. Ground vibration is the most hazardous issue of blasting which can cause critical damage to the surrounding structures. This paper focuses on developing an explicit model to predict the ground vibration through an multi objective evolutionary polynomial regression (MOGA-EPR). To this end, a database including 79 sets of data related to a quarry site in Malaysia were used. In addition, a gene expression programming (GEP) model and several empirical equations were employed to predict ground vibration, and their performances were then compared with the MOGA-EPR model using the mean absolute error (MAE), root mean square error (RMSE), mean (𝜇), standard deviation of the mean (𝜎), coefficient of determination (R²) and a20-index. Comparing the results, it was found that the MOGA-EPR model predicted the ground vibration more precisely than the GEP model and the empirical equations, where the MOGA-EPR scored lower MAE and RMSE, 𝜇 and 𝜎 closer to the optimum value, and higher R² and a20-index. Accordingly, the proposed MOGA-EPR model can be introduced as a useful method to predict ground vibration and has the capacity to be generalized to predict other blasting effects.

• Journal of radiological science and technology
• /
• v.46 no.2
• /
• pp.131-139
• /
• 2023

High-dose I-131 used for the treatment of thyroid cancer causes localized exposure among radiology technologists handling it. There is a delay between the calibration date and when the dose of I-131 is administered to a patient. Therefore, it is necessary to directly measure the radioactivity of the administered dose using a dose calibrator. In this study, we attempted to apply machine learning modeling to measured external dose rates from shielded I-131 in order to predict their radioactivity. External dose rates were measured at 1 m, 0.3 m, and 0.1 m distances from a shielded container with the I-131, with a total of 868 sets of measurements taken. For the modeling process, we utilized the hold-out method to partition the data with a 7:3 ratio (609 for the training set:259 for the test set). For the machine learning algorithms, we chose linear regression, decision tree, random forest and XGBoost. To evaluate the models, we calculated root mean square error (RMSE), mean square error (MSE), and mean absolute error (MAE) to evaluate accuracy and R² to evaluate explanatory power. Evaluation results are as follows. Linear regression (RMSE 268.15, MSE 71901.87, MAE 231.68, R² 0.92), decision tree (RMSE 108.89, MSE 11856.92, MAE 19.24, R² 0.99), random forest (RMSE 8.89, MSE 79.10, MAE 6.55, R² 0.99), XGBoost (RMSE 10.21, MSE 104.22, MAE 7.68, R² 0.99). The random forest model achieved the highest predictive ability. Improving the model's performance in the future is expected to contribute to lowering exposure among radiology technologists.

• Journal of Korean Association for Spatial Structures
• /
• v.24 no.3
• /
• pp.43-51
• /
• 2024

Dynamic responses of nuclear power plant structure subjected to earthquake loads should be carefully investigated for safety. Because nuclear power plant structure are usually constructed by material of reinforced concrete, the aging deterioration of R.C. have no small effect on structural behavior of nuclear power plant structure. Therefore, aging deterioration of R.C. nuclear power plant structure should be considered for exact prediction of seismic responses of the structure. In this study, a machine learning model for seismic response prediction of nuclear power plant structure was developed by considering aging deterioration. The OPR-1000 was selected as an example structure for numerical simulation. The OPR-1000 was originally designated as the Korean Standard Nuclear Power Plant (KSNP), and was re-designated as the OPR-1000 in 2005 for foreign sales. 500 artificial ground motions were generated based on site characteristics of Korea. Elastic modulus, damping ratio, poisson's ratio and density were selected to consider material property variation due to aging deterioration. Six machine learning algorithms such as, Decision Tree (DT), Random Forest (RF), Support Vector Machine (SVM), K-Nearest Neighbor (KNN), Artificial Neural Networks (ANN), eXtreme Gradient Boosting (XGBoost), were used t o construct seispic response prediction model. 13 intensity measures and 4 material properties were used input parameters of the training database. Performance evaluation was performed using metrics like root mean square error, mean square error, mean absolute error, and coefficient of determination. The optimization of hyperparameters was achieved through k-fold cross-validation and grid search techniques. The analysis results show that neural networks present good prediction performance considering aging deterioration.

• The Bulletin of The Korean Astronomical Society
• /
• v.45 no.1
• /
• pp.67.2-67.2
• /
• 2020

Not only the magnitude of the mirror surface error, the pattern matters as it produces certain aberrations. In particular, the surface error of the freeform mirrors, which are optimized to eliminate specific aberrations, might show much higher sensitivity in optical performance. Therefore, we analyze the mirror surface error with Zernike polynomials with the goal of generating a realistic error surface. We investigate the surface error of the freeform mirror fabricated by diamond turning machine to analyze the realistic tendency of the error. The surface error with 0.22 ㎛ root-mean-square value is fitted to the Zernike terms using the incremental fitting method, which increases the number of the fitting coefficients through steps. Furthermore, optical performance via surface error pattern based on Zernike terms is studied to see the influences of each term. With this study, realistic error surface generation may allow higher accuracy not only for the feasibility test but also for all tests and predictions using optical simulations.

• Journal of Biosystems Engineering
• /
• v.43 no.2
• /
• pp.148-159
• /
• 2018

Purpose: Unmanned air vehicle (UAV) remote sensing was applied to test various vegetation indices and make prediction models of protein content of rice for monitoring grain quality and proper management practice. Methods: Image acquisition was carried out by using NIR (Green, Red, NIR), RGB and RE (Blue, Green, Red-edge) camera mounted on UAV. Sampling was done synchronously at the geo-referenced points and GPS locations were recorded. Paddy samples were air-dried to 15% moisture content, and then dehulled and milled to 92% milling yield and measured the protein content by near-infrared spectroscopy. Results: Artificial neural network showed the better performance with $R^2$ (coefficient of determination) of 0.740, NSE (Nash-Sutcliffe model efficiency coefficient) of 0.733 and RMSE (root mean square error) of 0.187% considering all 54 samples than the models developed by PR (polynomial regression), SLR (simple linear regression), and PLSR (partial least square regression). PLSR calibration models showed almost similar result with PR as 0.663 ($R^2$) and 0.169% (RMSE) for cloud-free samples and 0.491 ($R^2$) and 0.217% (RMSE) for cloud-shadowed samples. However, the validation models performed poorly. This study revealed that there is a highly significant correlation between NDVI (normalized difference vegetation index) and protein content in rice. For the cloud-free samples, the SLR models showed $R^2=0.553$ and RMSE = 0.210%, and for cloud-shadowed samples showed 0.479 as $R^2$ and 0.225% as RMSE respectively. Conclusion: There is a significant correlation between spectral bands and grain protein content. Artificial neural networks have the strong advantages to fit the nonlinear problem when a sigmoid activation function is used in the hidden layer. Quantitatively, the neural network model obtained a higher precision result with a mean absolute relative error (MARE) of 2.18% and root mean square error (RMSE) of 0.187%.

• Computers and Concrete
• /
• v.31 no.3
• /
• pp.253-265
• /
• 2023

In this research, the gene expression programming (GEP) technique was employed to provide a new model for predicting the maximum loading capacity of concrete-encased steel (CES) columns. This model was developed based on 96 CES column specimens available in the literature. The six main parameters used in the model were the compressive strength of concrete (f_c), yield stress of structural steel (f_ys), yield stress of steel rebar (f_yr), and cross-sectional areas of concrete, structural steel, and steel rebar (A_c, A_s and A_r respectively). The performance of the prediction model for the ultimate load-carrying capacity was investigated using different statistical indicators such as root mean square error (RMSE), correlation coefficient (R), mean absolute error (MAE), and relative square error (RSE), the corresponding values of which for the proposed model were 620.28, 0.99, 411.8, and 0.01, respectively. Here, the predictions of the model and those of available codes including ACI ITG, AS 3600, CSA-A23, EN 1994, JGJ 138, and NZS 3101 were compared for further model assessment. The obtained results showed that the proposed model had the highest correlation with the experimental data and the lowest error. In addition, to see if the developed model matched engineering realities and corresponded to the previously developed models, a parametric study and sensitivity analysis were carried out. The sensitivity analysis results indicated that the concrete cross-sectional area (A_c) has the greatest effect on the model, while parameter (f_yr) has a negligible effect.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.7
• /
• pp.76-84
• /
• 2013

In the passive emitter localization using instantaneous TDOA (time difference of arrival) and FDOA (frequency difference of arrival) measurements, the estimation accuracy can be improved by collecting additional measurements. To achieve this goal, it is required to increase the number of the sensors. However, in electronic warfare environment, a large number of sensors cause the loss of military strength due to high probability of intercept. Also, the additional processes should be considered such as the data link and the clock synchronization between the sensors. Hence, in this paper, the passive localization of a stationary emitter is presented by using the successive TDOA and FDOA measurements from two moving sensors. In this case, since an independent pair of sensors is added in the data set at every instant of measurement, each pair of sensors does not share the common reference sensor. Therefore, the QCLS (quadratic correction least squares) methods cannot be applied, in which all pairs of sensor should include the common reference sensor. For this reason, a Gauss-Newton algorithm is adopted to solve the non-linear least square problem. In addition, to show the performance of the proposed method, we compare the RMSE (root mean square error) of the estimates with CRLB (Cramer-Rao lower bound) and derived the CEP (circular error probable) planes to analyze the expected estimation performance on the 2-dimensional space.

• Ocean and Polar Research
• /
• v.30 no.2
• /
• pp.129-140
• /
• 2008

As a preliminary effort to establish a data assimilative ocean forecasting system, we reviewed the theory of the Ensemble Kamlan Filter (EnKF) and developed practical techniques to apply the EnKF algorithm in a real ocean circulation modeling system. To verify the performance of the developed EnKF algorithm, a wind-driven double gyre was established in a rectangular ocean using the Regional Ocean Modeling System (ROMS) and the EnKF algorithm was implemented. In the ideal ocean, sea surface temperature and sea surface height were assimilated. The results showed that the multivariate background error covariance is useful in the EnKF system. We also tested the sensitivity of the EnKF algorithm to the localization and inflation of the background error covariance and the number of ensemble members. In the sensitivity tests, the ensemble spread as well as the root-mean square (RMS) error of the ensemble mean was assessed. The EnKF produces the optimal solution as the ensemble spread approaches the RMS error of the ensemble mean because the ensembles are well distributed so that they may include the true state. The localization and inflation of the background error covariance increased the ensemble spread while building up well-distributed ensembles. Without the localization of the background error covariance, the ensemble spread tended to decrease continuously over time. In addition, the ensemble spread is proportional to the number of ensemble members. However, it is difficult to increase the ensemble members because of the computational cost.

• Korean Journal of Remote Sensing
• /
• v.37 no.5_1
• /
• pp.861-870
• /
• 2021

In this study, fused Aerosol Optical Depth (AOD) data were produced using AOD products from the Geostationary Ocean Color Imager (GOCI) onboard Communication, Oceanography and Meteorology Satellite (COMS)satellite and the Advanced Himawari Imager (AHI) onboard Himawari-8. Since the spatial resolution and the coordinate system between the satellite sensors are different, a preprocessing was first preceded. After that, using the level 1.5 AOD dataset of AErosol RObotic NETwork (AERONET), which is ground-based observation, correlations and trends between each satellite AOD and AERONET AOD were utilized to produce more accurate satellite AOD data than the originalsatellite AODs. The fused AOD were found to be more accurate than the originalsatellite AODs. Root Mean Square Error (RMSE) and mean bias of the fused AODs were calculated to be 0.13 and 0.05, respectively. We also compared errors of the fused AODs against those of the original GOCI AOD (RMSE: 0.15, mean bias: 0.11) and the original AHI AOD (RMSE: 0.15, mean bias: 0.05). It was confirmed that the fused AODs have betterspatial coverage than the original AODsin areas where there are no observations due to the presence of cloud from a single satellite.