• Journal of the Korea institute for structural maintenance and inspection
• /
• v.23 no.6
• /
• pp.38-44
• /
• 2019

A Convolution Neural Network(CNN) model was utilized to detect surface cracks in asphalt concrete pavements. The CNN used for this study consists of five layers with 3×3 convolution filter and 2×2 pooling kernel. Pavement surface crack images collected by automated road surveying equipment was used for the training and testing of the CNN. The performance of the CNN was evaluated using the accuracy, precision, recall, missing rate, and over rate of the surface crack detection. The CNN trained with the largest amount of data shows more than 96.6% of the accuracy, precision, and recall as well as less than 3.4% of the missing rate and the over rate.

• The Journal of Information Systems
• /
• v.28 no.4
• /
• pp.333-358
• /
• 2019

Purpose The study aims to develop a data-based decision model for private bankers when recommending hedge funds to their customers in financial institutions. Design/methodology/approach The independent variables are set in two groups. The independent variables of the first group are aggressive investors, active investors, and risk-neutral type investors. In the second group, variables considered by private bankers include customer propensity to invest, reliability, product subscription experience, professionalism, intimacy, and product understanding. A decision-making variable for a private banker is in recommending a first-rate general private fund composed of foreign and domestic FinTech products. These contain dependent variables that include target return rate(%), fund period (months), safeguard existence, underlying asset, and hedge fund name. Findings Based on the research results, there is a 94.4% accuracy in decision-making when the independent variables (customer rating, reliability, intimacy, product subscription experience, professionalism and product understanding) are used according to the following order of relevant dependent variables: step 1 on safeguard existence, step 2 on target return rate, step 3 on fund period, and step 4 on hedge fund name. Next, a 93.7% accuracy is expected when decision-making uses the following order of dependent variables: step 1 on safeguard existence, step 2 on target return rate, step 3 on underlying asset, and step 4 on fund period. In conclusion, a private banker conducts a decision making stage when recommending hedge funds to their customers. When examining a private banker's recommendations of hedge funds to a customer, independent variables influencing dependent variables are intimacy, product comprehension, and product subscription experience according to a categorical regression model and artificial neural network analysis model.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.41 no.5
• /
• pp.391-399
• /
• 2017

This paper presents the effects of the cutting speed and feed rate on the axial shape of flat end-milled down cut side walls. Experiments were performed using the cutting speed, tool diameter, and feed per tooth as variables, and the thrust force and axial shape were measured as the experimental results. The results of this study confirmed that a smaller feed per tooth, which is proportional to the value obtained by dividing the feed rate by the cutting speed, results in a higher axial shape accuracy. In addition, the axial shape can be simplified to a form in which two straight lines having different slopes meet at a singular point. Therefore, it was concluded that the shape accuracy could easily be estimated during the operation and improved by adjusting the feed per tooth.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.17 no.12
• /
• pp.1391-1400
• /
• 1992

In this paper, we suggested a method for evaluating the type-I/II error which is proposed by the CCITT as a criterion for the accuracy of policing algorithms in ATM networks, By the analysis of the type-I/II error of the Leaky Bucket(LB) algorithm, we investigated the relationships between the traffic parameters and the LB parameters to police the mean and peak cell rate effectively in the ON/OFF traffic. We showed that the LB parameters, the leaky rate a and the threshold M of the LB counter, could be determined as a pair of (a, M) satisfying the type-I/II error and minimizing the response time. In the ON/OFF traffic, it has been observed that the a-M characteristic curve of the LB policing algorithm only depends on the burstiness. As the results of the performance analysis, we found that the LB algorithm exhibits a good performance in the peak rate policing, but has some problems in the mean rate policing due to the trade-off between the accuracy and the response time.

• Journal of Biomedical Engineering Research
• /
• v.37 no.2
• /
• pp.61-67
• /
• 2016

This paper presented a method for random forest based the arrhythmia classification using both heart rate (HR) and heart rate variability (HRV) features. We analyzed the MIT-BIH arrhythmia database which contains half-hour ECG recorded from 48 subjects. This study included not only the linear features but also non-linear features for the improvement of classification performance. We classified abnormal ECG using mean_NN (mean of heart rate), SD1/SD2 (geometrical feature of poincare HRV plot), SE (spectral entropy), pNN100 (percentage of a heart rate longer than 100 ms) affecting accurate classification among combined of linear and nonlinear features. We compared our proposed method with Neural Networks to evaluate the accuracy of the algorithm. When we used the features extracted from the HRV as an input variable for classifier, random forest used only the most contributed variable for classification unlike the neural networks. The characteristics of random forest enable the dimensionality reduction of the input variables, increase a efficiency of classifier and can be obtained faster, 11.1% higher accuracy than the neural networks.

• The Korean Journal of Applied Statistics
• /
• v.30 no.6
• /
• pp.993-1004
• /
• 2017

A stratified sampling method is generally used with a sample selected using the same sample weight in each stratum in order to improve the accuracy of the sampling survey estimation. However, the weight should be adjusted to reflect the response rate if the response rate is affected by the value of the variable of interest. It may be also more effective to adjust the weights by subdividing the stratum rather than using the same weight if the variable of interest has a linear relationship with the continuous auxiliary variables. In this study, we propose a method to increase the accuracy of estimation using an informative sampling design technique when the response rate is an exponential function of the variable of interest and the variable of interest has a linear relationship with the auxiliary variable. Simulation results show the superiority of the proposed method.

• Journal of the Korean Society of Manufacturing Technology Engineers
• /
• v.25 no.2
• /
• pp.124-131
• /
• 2016

Several research works for finding and optimizing the methods of dispensing high viscosity phosphor used in the fabrication of white LED's are currently in progress. High viscosity phosphor dispensing with a high accuracy is crucial because the dispensing rate and uniformity directly affect parameters such as the CIE chromaticity diagram, color temperature and luminous flux of white LED's. This study presents a novel method of dispensing high viscosity phosphor using electrohydrodynamic printing. The dispensing rate was optimized less than 0.01 mg phosphor using experiments and optimizing the process parameters including the standoff distance from the nozzle to the substrate, ink supply pressure, and multi-step pulsed waveform magnitude ratio. The dispensing rate was measured by dispensing 20 dots using drop-on-demand with the optimized parameters, and the experiments were repeated 10 times to maximize the data accuracy. The average dispensing rate that can be reliably used for high viscosity phosphor dispensing was 0.0052 mg.

• Structural Engineering and Mechanics
• /
• v.73 no.4
• /
• pp.427-436
• /
• 2020

The use of final IGS precise orbit and clock products for high-rate GNSS-PPP proved its effectiveness in capturing dynamic displacement of engineering structures caused by earthquakes. However, the main drawback of using the final products is that they are available after approximately two weeks of data collection, which is not suitable for timely measures after an event. In this study, the use of ultra-rapid products (observed part), which are available after a few hours of data collection, and rapid products, which are available in less than 24 hrs, are investigated and their results are compared to the more precise final products. The tests are designed such that harmonic oscillations with different frequencies and amplitudes and ground motion of a simulated real earthquake are generated using a single axis shake table and the PPP was used to capture these movements by monitoring time-change of the table positions. To evaluate the accuracy of PPP using ultra-rapid, rapid and final products, their results were compared with relative GNSS positioning and LVDT (Linear Variable Differential Transformer) data, treated as reference. The results show that the high-rate GNSS-PPP solutions based on the three products can capture frequencies of harmonic oscillations and dynamic displacement with good accuracy. There were slight differences between ultra-rapid, rapid and final products, where some of the tested events indicated that the latter two produced are more accurate and provide better results compared to the ultra-rapid product for monitoring short-term dynamic displacements.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.25 no.4
• /
• pp.476-481
• /
• 2019

Early detection of fire is an important measure for minimizing the loss of life and property damage. However, fire and smoke need to be simultaneously detected. In this context, numerous studies have been conducted on image-based fire detection. Conventional fire detection methods are compute-intensive and comprise several algorithms for extracting the flame and smoke characteristics. Hence, deep learning algorithms and convolution neural networks can be alternatively employed for fire detection. In this study, recorded image data of fire in a ship engine room were analyzed. The flame and smoke characteristics were extracted from the outer box, and the YOLO (You Only Look Once) convolutional neural network algorithm was subsequently employed for learning and testing. Experimental results were evaluated with respect to three attributes, namely detection rate, error rate, and accuracy. The respective values of detection rate, error rate, and accuracy are found to be 0.994, 0.011, and 0.998 for the flame, 0.978, 0.021, and 0.978 for the smoke, and the calculation time is found to be 0.009 s.

• Journal of the Korea Institute of Building Construction
• /
• v.19 no.5
• /
• pp.449-456
• /
• 2019

Recent developments in architectural technologies and programs have enabled architects to think creatively and design free-form architecture. however, there are many problems in the production technology of FCP(Free-Form Concrete Panel). In particular, reduced accuracy due to lack of free-form panel production technology can lead to redesign of buildings as a result, problems such as an increase in construction cost and period. Therefore, this experiment aimed to compensate the decrease of the accuracy according to the displacement difference and to derive the proper mixing ratio for maintaining the shape during the free-form panel curing. In this study, molds were made using paraffin that is a recyclable phase change material. Concrete Panel is usually produced from Portland cement, dead burn magnesia, phosphate, borax and fine aggregate. In this study, four mixing ratios of FCP were selected after each material was blended to determine the proper blending ratio of the fluidity phase, the water absorption rate and the water content of the test piece. FCP was fabricated on the basis of the selected four compounding ratios and thickness and error rate were measured. Based on the error rate of the measured FCP, the quality standard was satisfied among the four compounding ratios.