• Journal of the Korean Institute of Intelligent Systems
• /
• v.17 no.4
• /
• pp.435-442
• /
• 2007

To reduce damage from high concentration ozone in the air, we have researched how to predict high concentration ozone before it occurs. High concentration ozone is a rare event and its reaction mechanism has nonlinearities and complexities. In this paper, we have tried to apply and consider as many methods as we could. We clustered the data using the fuzzy c-mean method and took a rejection sampling to fill in the missing and abnormal data. Next, correlations of the input component and output ozone concentration were calculated to transform more correlated components by modified log transformation. Then, we made the prediction models using Dynamic Polynomial Neural Networks. To select the optimal model, we adopted a minimum bias criterion. Finally, to evaluate suggested models, we compared the two models. One model was trained and tested by the transformed data and the other was not. We concluded that the modified transformation effected good to ideal performance In some evaluations. In particular, the data were related to seasonal characteristics or its variation trends.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.16 no.5
• /
• pp.608-615
• /
• 2006

This paper proposes a fuzzy model tree consisting of local linear models using fuzzy cluster for data modeling. First, cluster centers are calculated by fuzzy clustering method using all input and output attributes. And then, linear models are constructed at internal nodes with fuzzy membership values between centers and input attributes. The expansion of internal node is determined by comparing errors calculated in parent node with ones in child node, respectively. As a final step, data prediction is performed with a linear model having the highest fuzzy membership value between input attributes and cluster centers in leaf nodes. To show the effectiveness of the proposed method, we have applied our method to various dataset. Under various experiments, our proposed method shows better performance than conventional model tree and artificial neural networks.

• Journal of Digital Convergence
• /
• v.19 no.9
• /
• pp.181-188
• /
• 2021

Research and importance of legal services applied with AI so that it can be easily understood and predictable in difficult legal fields is increasing. In this study, based on the decision of the Tax Tribunal in the field of tax law, a model was built through self-learning through information collection and data processing, and the prediction results were answered to the user's query and the accuracy was verified. The proposed model collects information on tax decisions and extracts useful data through web crawling, and generates word vectors by applying Word2Vec's Fast Text algorithm to the optimized output through NLP. 11,103 cases of information were collected and classified from 2017 to 2019, and verified with 70% accuracy. It can be useful in various legal systems and prior research to be more efficient application.

• Atmosphere
• /
• v.31 no.1
• /
• pp.1-15
• /
• 2021

High-resolution wind resources maps (maps, here after) with spatial and temporal resolutions of 100 m and 3-hours, respectively, over South Korea have been produced and evaluated for the period from July 2016 to June 2017 using Korea Meteorological Administration (KMA) Post Processing (KMAPP). Evaluation of the 10 m- and 80 m-level wind speed in the new maps (KMAPP-Wind) and the 1.5 km-resolution KMA NWP model, Local Data Assimilation and Prediction System (LDAPS), shows that the new high-resolution maps improves of the LDAPS winds in estimating the 10m wind speed as the new data reduces the mean bias (MBE) and root-mean-square error (RMSE) by 33.3% and 14.3%, respectively. In particular, the result of evaluation of the wind at 80 m which is directly related with power turbine shows that the new maps has significantly smaller error compared to the LDAPS wind. Analyses of the new maps for the seasonal average, maximum wind speed, and the prevailing wind direction shows that the wind resources over South Korea are most abundant during winter, and that the prevailing wind direction is strongly affected by synoptic weather systems except over mountainous regions. Wind speed generally increases with altitude and the proximity to the coast. In conclusion, the evaluation results show that the new maps provides significantly more accurate wind speeds than the lower resolution NWP model output, especially over complex terrains, coastal areas, and the Jeju island where wind-energy resources are most abundant.

• Journal of the Architectural Institute of Korea Structure & Construction
• /
• v.34 no.6
• /
• pp.63-69
• /
• 2018

The machine learning model can capture the dynamics of building systems with less inputs than the first principle based simulation model. The training data for developing a machine learning model are usually selected in a heuristic manner. In this study, the authors developed a machine learning model which can describe supply air temperature from an AHU in a real office building. For rational reduction of the training data, the progressive sampling method was used. It is found that even though the progressive sampling requires far less training data (n=60) than the offline regular sampling (n=1,799), the MBEs of both models are similar (2.6% vs. 5.4%). In addition, for the update of the machine learning model, the normalized mutual information (NMI) was applied. If the NMI between the simulation output and the measured data is less than 0.2, the model has to be updated. By the use of the NMI, the model can perform better prediction ($5.4%{\rightarrow}1.3%$).

• Journal of The Korean Society of Agricultural Engineers
• /
• v.64 no.1
• /
• pp.79-89
• /
• 2022

Soil moisture is significantly related to crop growth and plays an important role in irrigation management. To predict soil moisture, various process-based model has been developed and used in the world. Various models (Land surface model) may have different performance depending on the model parameters and structures that causes the different model output for the same modeling condition. In this study, the three land surface models (Noah Land Surface Model, Soil Water Atmosphere Plant, Community Land Model) were used to compare the model performance (soil moisture prediction) and develop the multi-model simulation. At first, the genetic algorithm was used to estimate the optimal soil parameters for each model, and the parameters were used to predict soil moisture in the study area. Then, we used the multi-model approach based on Bayesian model averaging (BMA). The results derived from this approach showed a better match to the measurements than the results from the original single land surface model. In addition, identifying the strengths and weaknesses of the single model and utilizing multi-model methods can help to increase the accuracy of soil moisture prediction.

• Journal of Aerospace System Engineering
• /
• v.16 no.4
• /
• pp.17-27
• /
• 2022

In this study, a deep learning-based network that can predict the aerodynamic characteristics of airfoils was designed, and the feasibility of the proposed network was confirmed by applying aerodynamic data generated by Xfoil. The prediction of aerodynamic characteristics according to the variation of airfoil thickness was performed. Considering the angle of attack, the coordinate data of an airfoil is converted into image data using signed distance function. Additionally, the distribution of the pressure coefficient on airfoil is expressed as reduced data via proper orthogonal decomposition, and it was used as the output of the proposed network. The test data were constructed to evaluate the interpolation and extrapolation performance of the proposed network. As a result, the coefficients of determination of the lift coefficient and moment coefficient were confirmed, and it was found that the proposed network shows benign performance for the interpolation test data, when compared to that of the extrapolation test data.

• Imaging Science in Dentistry
• /
• v.52 no.3
• /
• pp.239-244
• /
• 2022

Purpose: Despite the proliferation of numerous morphometric and anthropometric methods for sex identification based on linear, angular, and regional measurements of various parts of the body, these methods are subject to error due to the observer's knowledge and expertise. This study aimed to explore the possibility of automated sex determination using convolutional neural networks(CNNs) based on lateral cephalometric radiographs. Materials and Methods: Lateral cephalometric radiographs of 1,476 Iranian subjects (794 women and 682 men) from 18 to 49 years of age were included. Lateral cephalometric radiographs were considered as a network input and output layer including 2 classes(male and female). Eighty percent of the data was used as a training set and the rest as a test set. Hyperparameter tuning of each network was done after preprocessing and data augmentation steps. The predictive performance of different architectures (DenseNet, ResNet, and VGG) was evaluated based on their accuracy in test sets. Results: The CNN based on the DenseNet121 architecture, with an overall accuracy of 90%, had the best predictive power in sex determination. The prediction accuracy of this model was almost equal for men and women. Furthermore, with all architectures, the use of transfer learning improved predictive performance. Conclusion: The results confirmed that a CNN could predict a person's sex with high accuracy. This prediction was independent of human bias because feature extraction was done automatically. However, for more accurate sex determination on a wider scale, further studies with larger sample sizes are desirable.

• Geomechanics and Engineering
• /
• v.33 no.3
• /
• pp.279-289
• /
• 2023

Brittleness index (BI) is an important property of rocks because it is a good index to predict rockburst. Due to its importance, several empirical and soft computing (SC) models have been proposed in the literature based on the punch penetration test (PPT) results. These models are very important as there is no clear-cut experimental means for measuring BI asides the PPT which is very costly and time consuming to perform. This study used a novel Multivariate Adaptive regression spline (MARS), M5P, and white-box ANN to predict the BI of rocks using the available data in the literature for an improved BI prediction. The rock density, uniaxial compressive strength (σ_c) and tensile strength (σ_t) were used as the input parameters into the models while the BI was the targeted output. The models were implemented in the MATLAB software. The results of the proposed models were compared with those from existing multilinear regression, linear and nonlinear particle swarm optimization (PSO) and genetic algorithm (GA) based models using similar datasets. The coefficient of determination (R²), adjusted R² (Adj R²), root-mean squared error (RMSE) and mean absolute percentage error (MAPE) were the indices used for the comparison. The outcomes of the comparison revealed that the proposed ANN and MARS models performed better than the other models with R² and Adj R² values above 0.9 and least error values while the M5P gave similar performance to those of the existing models. Weight partitioning method was also used to examine the percentage contribution of model predictors to the predicted BI and tensile strength was found to have the highest influence on the predicted BI.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.44 no.6
• /
• pp.57-65
• /
• 2007

This paper presents the design of single phase PFC(Power Factor Correction) using a DSP(TMS320F2812). In order to realize the proposed boost PFC converter in average current mode control, the DSP requires the A/D sampling values for a line input voltage, a inductor current, and the output voltage of the converter. Because of a FET switching noise, these sampling values contain a high frequency noise and switching ripple. The solution of A/D sampling keeps away from the switching point. Because the PWM duty is changed from 5% to 95%, we can#t decide a fixed sampling time. In this paper, the three A/D converters of the DSP are started using the prediction algorithm for the FET ON/OFF time at every sampling cycle(40 KHz). Implemented A/D sampling algorithm with only one timer of the DSP is very simple and gives the autostart of these A/D converters. From the experimental result, it was shown that the power factor was about 0.99 at wide input voltage, and the output ripple voltage was smaller than 5 Vpp at 80 Vdc output. Finally the parameters and gains of PI controllers are controlled by serial communication with Windows Xp based PC. Also it was shown that the implemented PFC converter can achieve the feasibility and the usefulness.