• Journal of Korea Water Resources Association
• /
• v.55 no.11
• /
• pp.903-911
• /
• 2022

Groundwater, one of the resources for supplying water, fluctuates in water level due to various natural factors. Recently, research has been conducted to predict fluctuations in groundwater levels using Artificial Neural Network (ANN). Previously, among operators in ANN, Gradient Descent (GD)-based Optimizers were used as Optimizer that affect learning. GD-based Optimizers have disadvantages of initial correlation dependence and absence of solution comparison and storage structure. This study developed Gradient Descent combined with Harmony Search (GDHS), a new Optimizer that combined GD and Harmony Search (HS) to improve the shortcomings of GD-based Optimizers. To evaluate the performance of GDHS, groundwater level at Icheon Yullhyeon observation station were learned and predicted using Multi Layer Perceptron (MLP). Mean Squared Error (MSE) and Mean Absolute Error (MAE) were used to compare the performance of MLP using GD and GDHS. Comparing the learning results, GDHS had lower maximum, minimum, average and Standard Deviation (SD) of MSE than GD. Comparing the prediction results, GDHS was evaluated to have a lower error in all of the evaluation index than GD.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.39 no.6
• /
• pp.351-361
• /
• 2021

The seriousness of fire is rising because fire causes enormous damage to property and human life. Therefore, this study aims to predict various risk factors affecting fire by fire type. The predictive analysis of fire factors was carried out targeting Gyeonggi-do, which has the highest number of fires in the country. For the analysis, using machine learning methods SVM (Support Vector Machine), RF (Random Forest), GBRT (Gradient Boosted Regression Tree) the accuracy of each model was presented with a high fit model through MAE (Mean Absolute Error) and RMSE (Root Mean Squared Error), and based on this, predictive analysis of fire factors in Gyeonggi-do was conducted. In addition, using machine learning methods such as SVM (Support Vector Machine), RF (Random Forest), and GBRT (Gradient Boosted Regression Tree), the accuracy of each model was presented with a high-fit model through MAE and RMSE. Predictive analysis of occurrence factors was achieved. Based on this, as a result of comparative analysis of three machine learning methods, the RF method showed a MAE = 1.765 and RMSE = 1.876, as well as the MAE and RMSE verification and test data were very similar with a difference between MAE = 0.046 and RMSE = 0.04 showing the best predictive results. The results of this study are expected to be used as useful data for fire safety management allowing decision makers to identify the sequence of dangers related to the factors affecting the occurrence of fire.

• Imaging Science in Dentistry
• /
• v.49 no.1
• /
• pp.19-26
• /
• 2019

Purpose: It has been proposed that using new prediction methods, such as neural networks based on dental data, could improve age estimation. This study aimed to assess the possibility of exploiting neural networks for estimating age by means of the pulp-to-tooth ratio in canines as a non-destructive, non-expensive, and accurate method. In addition, the predictive performance of neural networks was compared with that of a linear regression model. Materials and Methods: Three hundred subjects whose age ranged from 14 to 60 years and were well distributed among various age groups were included in the study. Two statistical software programs, SPSS 21 (IBM Corp., Armonk, NY, USA) and R, were used for statistical analyses. Results: The results indicated that the neural network model generally performed better than the regression model for estimation of age with pulp-to-tooth ratio data. The prediction errors of the developed neural network model were acceptable, with a root mean square error (RMSE) of 4.40 years and a mean absolute error (MAE) of 4.12 years for the unseen dataset. The prediction errors of the regression model were higher than those of the neural network, with an RMSE of 10.26 years and a MAE of 8.17 years for the test dataset. Conclusion: The neural network method showed relatively acceptable performance, with an MAE of 4.12 years. The application of neural networks creates new opportunities to obtain more accurate estimations of age in forensic research.

• Journal of Internet Computing and Services
• /
• v.8 no.6
• /
• pp.95-102
• /
• 2007

Recommender System based on the Collaborative Filtering has a problem of trust of the prediction accuracy because of its problem of sparsity. If the sparsity of a preference value is large, it causes a problem on a process of a choice of neighbors and also lowers the prediction accuracy. In this article, a change of MAE based on the sparsity is studied, groups are classified by sparsity and then, the significant difference among MAEs of classified groups is analyzed. To improve the accuracy of prediction among groups by the problem of sparsity, We studied the improvement of an accurate prediction for recommending system through reducing sparsity by sorting sparsity items, and replacing the average preference among them that has a lot of respondents with the preference evaluation value.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.54 no.2
• /
• pp.123-129
• /
• 2017

This paper proposes a new nonnegative matrix factorization (NMF) based direction-of-arrival (DOA) estimation method for multiple sound sources using a dual microphone array. First of all, sound signals coming from the dual microphone array are segmented into consecutive analysis frames, and a steered-response power phase transform (SRP-PHAT) beamformer is applied to each frame so that stereo signals of each frame are represented in a time-direction domain. The time-direction outputs of SRP-PHAT are stored for a pre-defined number of frames, which is referred to as a time-direction block. Next, In order to estimate DOAs robust to noise, each time-direction block is normalized along the time by using a block subtraction technique. After that, an unsupervised NMF method is applied to the normalized time-direction block in order to cluster the directions of each sound source in a multiple sound source environments. In particular, the activation and basis matrices are used to estimate the number of sound sources and their DOAs, respectively. The DOA estimation performance of the proposed method is evaluated by measuring a mean absolute error (MAE) and the standard deviation of errors between the oracle and estimated DOAs under a three source condition, where the sources are located in [$-35{\circ}$, 5m], [$12{\circ}$, 4m], and [$38{\circ}$, 4.m] from the dual microphone array. It is shown from the experiment that the proposed method could relatively reduce MAE by 56.83%, compared to a conventional SRP-PHAT based DOA estimation method.

• Korean Journal of Remote Sensing
• /
• v.37 no.5_3
• /
• pp.1405-1423
• /
• 2021

Precipitation is one of the main factors that affect water and energy cycles, and its estimation plays a very important role in securing water resources and timely responding to water disasters. Satellite-based quantitative precipitation estimation (QPE) has the advantage of covering large areas at high spatiotemporal resolution. In this study, machine learning-based rainfall intensity models were developed using Himawari-8 Advanced Himawari Imager (AHI) water vapor channel (6.7 ㎛), infrared channel (10.8 ㎛), and weather radar Column Max (CMAX) composite data based on random forest (RF). The target variables were weather radar reflectivity (dBZ) and rainfall intensity (mm/hr) converted by the Z-R relationship. The results showed that the model which learned CMAX reflectivity produced the Critical Success Index (CSI) of 0.34 and the Mean-Absolute-Error (MAE) of 4.82 mm/hr. When compared to the GeoKompsat-2 and Precipitation Estimation from Remotely Sensed Information Using Artificial Neural Networks (PERSIANN)-Cloud Classification System (CCS) rainfall intensity products, the accuracies improved by 21.73% and 10.81% for CSI, and 31.33% and 23.49% for MAE, respectively. The spatial distribution of the estimated rainfall intensity was much more similar to the radar data than the existing products.

• Korean Chemical Engineering Research
• /
• v.61 no.1
• /
• pp.123-141
• /
• 2023

A sufficient amount of data with quality is needed for training artificial neural networks (ANNs). However, developing ANN models with a small amount of data often appears in engineering fields. This paper presented an ANN model to improve prediction performance of the ammonia emission amount with 83 data. The ammonia emission rate included eleven inputs and two outputs (maximum ammonia loss, N_max and time to reach half of N_max, K_m). Categorical input variables were transformed into multi-dimensional equal-distance variables, and 13 data were added into 66 training data using a generative adversarial network. Hyperparameters (number of layers, number of neurons, and activation function) of ANN were optimized using Gaussian process. Using 17 test data, the previous ANN model (Lim et al., 2007) showed the mean absolute error (MAE) of K_m and N_max to 0.0668 and 0.1860, respectively. The present ANN outperformed the previous model, reducing MAE by 38% and 56%.

• Korean Journal of Mineralogy and Petrology
• /
• v.37 no.2
• /
• pp.23-34
• /
• 2024

It is essential to know the mineral composition of core samples to assess the possibility of gas hydrate (GH) in sediments. During the exploration of gas hydrates (GH), mineral composition values were obtained from each core sample collected in the Ulleung Basin using X-ray diffraction (XRD). Based on this data, machine learning was performed with 3100 input values representing XRD peak intensities and 12 output values representing mineral compositions. The 488 data points were divided into 307 training samples, 132 validation samples, and 49 test samples. The random forest (RF) algorithm was utilized to obtain results. The machine learning results, compared with expert-predicted mineral compositions, revealed a Mean Absolute Error (MAE) of 1.35%. To enhance the performance of the developed model, principal component analysis (PCA) was employed to extract the key features of XRD peaks, reducing the dimensionality of input data. Subsequent machine learning with the refined data resulted in a decreased MAE, reaching a maximum of 1.23%. Additionally, the efficiency of the learning process improved over time, as confirmed from a temporal perspective.

• Journal of Electrical Engineering and Technology
• /
• v.10 no.5
• /
• pp.2197-2204
• /
• 2015

In recent years, medical image diagnosis has growing significant momentous in the medicinal field. Brain and lung image of patient are distorted with salt and pepper noise is caused by moving the head and chest during scanning process of patients. Reconstruction of these images is a most significant field of diagnostic evaluation and is produced clearly through techniques such as linear or non-linear filtering. However, restored images are produced with smaller amount of noise reduction in the presence of huge magnitude of salt and pepper noises. To eliminate the high density of salt and pepper noises from the reproduction of images, a new efficient fuzzy based median filtering algorithm with a moderate elapsed time is proposed in this paper. Reproduction image results show enhanced performance for the proposed algorithm over other available noise reduction filtering techniques in terms of peak signal -to -noise ratio (PSNR), mean square error (MSE), root mean square error (RMSE), mean absolute error (MAE), image enhancement factor (IMF) and structural similarity (SSIM) value when tested on different medical images like magnetic resonance imaging (MRI) and computer tomography (CT) scan brain image and CT scan lung image. The introduced algorithm is switching filter that recognize the noise pixels and then corrects them by using median filter with fuzzy two-sided π- membership function for extracting the local information.

• Journal of Korean Association for Spatial Structures
• /
• v.23 no.3
• /
• pp.87-94
• /
• 2023

Chloride is one of the most common threats to reinforced concrete (RC) durability. Alkaline environment of concrete makes a passive layer on the surface of reinforcement bars that prevents the bar from corrosion. However, when the chloride concentration amount at the reinforcement bar reaches a certain level, deterioration of the passive protection layer occurs, causing corrosion and ultimately reducing the structure's safety and durability. Therefore, understanding the chloride diffusion and its prediction are important to evaluate the safety and durability of RC structure. In this study, the chloride diffusion coefficient is predicted by machine learning techniques. Various machine learning techniques such as multiple linear regression, decision tree, random forest, support vector machine, artificial neural networks, extreme gradient boosting annd k-nearest neighbor were used and accuracy of there models were compared. In order to evaluate the accuracy, root mean square error (RMSE), mean square error (MSE), mean absolute error (MAE) and coefficient of determination (R2) were used as prediction performance indices. The k-fold cross-validation procedure was used to estimate the performance of machine learning models when making predictions on data not used during training. Grid search was applied to hyperparameter optimization. It has been shown from numerical simulation that ensemble learning methods such as random forest and extreme gradient boosting successfully predicted the chloride diffusion coefficient and artificial neural networks also provided accurate result.