• Information Systems Review
• /
• v.22 no.2
• /
• pp.201-217
• /
• 2020

Information security is an essential element not only to ensure the operation of the company and trust with customers but also to mitigate uncertain damage by preventing information data breach. Therefore, It is important to select appropriate information security countermeasures and determine the appropriate level of investment. This study presents a decision support model for the appropriate investment amount for each countermeasure as well as an optimal portfolio of information countermeasures within a limited budget. We analyze statistics on the types of information security breach by industry and derive an optimal portfolio of information security countermeasures by using genetic algorithms. The results of this study suggest guidelines for investing in information security countermeasures in various industries and help to support objective information security investment decisions.

• Journal of Animal Reproduction and Biotechnology
• /
• v.39 no.2
• /
• pp.138-144
• /
• 2024

Background: As the number of households raising companion dogs increases, the pet genetic analysis market also continues to grow. However, most studies have focused on specific purposes or native breeds. This study aimed to collect genomic data through single nucleotide polymorphism (SNP) chip analysis of companion dogs in South Korea and perform genetic diversity analysis and SNP annotation. Methods: We collected samples from 95 dogs belonging to 26 breeds, including mixed breeds, in South Korea. The SNP genotypes were obtained for each sample using an Axiom^TM Canine HD Array. Quality control (QC) was performed to enhance the accuracy of the analysis. A genetic diversity analysis was performed for each SNP. Results: QC initially selected SNPs, and after excluding non-diverse ones, 621,672 SNPs were identified. Genetic diversity analysis revealed minor allele frequencies, polymorphism information content, expected heterozygosity, and observed heterozygosity values of 0.220, 0.244, 0.301, and 0.261, respectively. The SNP annotation indicated that most variations had an uncertain or minimal impact on gene function. However, approximately 16,000 non-synonymous SNPs (nsSNPs) have been found to significantly alter gene function or affect exons by changing translated amino acids. Conclusions: This study obtained data on SNP genetic diversity and functional SNPs in companion dogs raised in South Korea. The results suggest that establishing an SNP set for individual identification could enable a gene-based registration system. Furthermore, identifying and researching nsSNPs related to behavior and diseases could improve dog care and prevent abandonment.

• Journal of Wind Energy
• /
• v.13 no.3
• /
• pp.13-21
• /
• 2022

When a wind turbine is designed, the dynamic stability of the system as well as the dynamic characteristics of the main components such as blades, hub, main shaft and tower must be evaluated. In particular, the natural frequencies of a blade, as a main load-generating component, need to be measured and assessed by component level testing. In conventional practice, the natural frequencies of a blade are determined as the measured frequencies near the reference frequencies provided by FE analysis results. But the reference frequencies are also uncertain since designers have difficulty distinguishing the torsional mode shape among the analysis results due to the complexity of its mode shape. So, in conventional practice, the determination of a measured torsional frequency inevitably contains uncertainty. Therefore, a novel method to definitely determine the torsional frequencies from the experimental data itself is necessary. In this paper, a new methodology to measure the torsional frequency of a blade was studied from the perspective of a modal test procedure, data processing method and mode determination logic. Finally, the validity of the method that can measure torsional frequency without reference FE analysis results was verified by applying it to an actual large wind turbine blade

• Korean Journal of Remote Sensing
• /
• v.36 no.5_3
• /
• pp.1095-1107
• /
• 2020

Flood monitoring using satellite data has been constrained by obtaining satellite images for flood peak and accurately extracting flooded areas from satellite data. Deep learning is a promising method for satellite image classification, yet the potential of deep learning-based flooded area extraction using SAR data remained uncertain, which has advantages in obtaining data, comparing to optical satellite data. This research explores the performance of SegNet and U-Net on image segmentation by extracting flooded areas in the Khorat basin, Mekong river basin, and Cagayan river basin in Thailand, Laos, and the Philippines from Sentinel-1 A/B satellite data. Results show that Global Accuracy, Mean IoU, and Mean BF Score of SegNet are 0.9847, 0.6016, and 0.6467 respectively, whereas those of U-Net are 0.9937, 0.7022, 0.7125. Visual interpretation shows that the classification accuracy of U-Net is higher than SegNet, but overall processing time of SegNet is around three times faster than that of U-Net. It is anticipated that the results of this research could be used when developing deep learning-based flood monitoring models and presenting fully automated flooded area extraction models.

• Journal of Internet Computing and Services
• /
• v.24 no.5
• /
• pp.17-27
• /
• 2023

Today, as AI (Artificial Intelligence) technology develops and its practicality increases, it is widely used in various application fields in real life. At this time, the AI model is basically learned based on various statistical properties of the learning data and then distributed to the system, but unexpected changes in the data in a rapidly changing data situation cause a decrease in the model's performance. In particular, as it becomes important to find drift signals of deployed models in order to respond to new and unknown attacks that are constantly created in the security field, the need for lifecycle management of the entire model is gradually emerging. In general, it can be detected through performance changes in the model's accuracy and error rate (loss), but there are limitations in the usage environment in that an actual label for the model prediction result is required, and the detection of the point where the actual drift occurs is uncertain. there is. This is because the model's error rate is greatly influenced by various external environmental factors, model selection and parameter settings, and new input data, so it is necessary to precisely determine when actual drift in the data occurs based only on the corresponding value. There are limits to this. Therefore, this paper proposes a method to detect when actual drift occurs through an Anomaly analysis technique based on XAI (eXplainable Artificial Intelligence). As a result of testing a classification model that detects DGA (Domain Generation Algorithm), anomaly scores were extracted through the SHAP(Shapley Additive exPlanations) Value of the data after distribution, and as a result, it was confirmed that efficient drift point detection was possible.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.39 no.8
• /
• pp.751-758
• /
• 2015

Mathematical models are actively used to reduce the experimental expenses required to understand physical phenomena. However, they are different from real phenomena because of assumptions or uncertain parameters. In this study, we present a calibration and validation method using a paper helicopter and statistical methods to quantify the uncertainty. The data from the experiment using three nominally identical paper helicopters consist of different groups, and are used to calibrate the drag coefficient, which is an unknown input parameter in both analytical models. We predict the predicted fall time data using probability distributions. We validate the analysis models by comparing the predicted distribution and the experimental data distribution. Moreover, we quantify the uncertainty using the Markov Chain Monte Carlo method. In addition, we compare the manufacturing error and experimental error obtained from the fall-time data using Analysis of Variance. As a result, all of the paper helicopters are treated as one identical model.

• Journal of Digital Convergence
• /
• v.12 no.5
• /
• pp.139-148
• /
• 2014

In recent years, computational models using parametric estimation method have been developed and used widely for efficient cost analysis. In this research, by applying experienced data from Guidance and Control Systems in Missile System field, the cost analysis for engineering model and commercial computational model(Price H, HL, M, S) are conducted and its result is analysed, so that the difference between two models and its grounds are apprehended. Comparing the calibrated value of computational model based on the data base of similar equipment and the cost from the engineering estimation, the two results are very close. It means that the credibility of data is enhanced through calibration. Also, for cost analysis of similar components in the future, the method for calibration of the computational models is also examined. When estimating development cost in this research, although many parts have been estimated through uncertain elements, the reliability could have been enhanced by applying computational model which secures objectivity. It is a very reasonable estimation method by utilizing calibration of the computational models based on existing accumulated development data.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.2
• /
• pp.59-76
• /
• 2023

Since the COVID-19 era, the rise in apartment prices has been unconventional. In this uncertain real estate market, price prediction research is very important. In this paper, a model is created to predict the actual transaction price of future apartments after building a vast data set of 870,000 from 2015 to 2020 through data collection and crawling on various real estate sites and collecting as many variables as possible. This study first solved the multicollinearity problem by removing and combining variables. After that, a total of five variable selection algorithms were used to extract meaningful independent variables, such as Forward Selection, Backward Elimination, Stepwise Selection, L1 Regulation, and Principal Component Analysis(PCA). In addition, a total of four machine learning and deep learning algorithms were used for deep neural network(DNN), XGBoost, CatBoost, and Linear Regression to learn the model after hyperparameter optimization and compare predictive power between models. In the additional experiment, the experiment was conducted while changing the number of nodes and layers of the DNN to find the most appropriate number of nodes and layers. In conclusion, as a model with the best performance, the actual transaction price of apartments in 2021 was predicted and compared with the actual data in 2021. Through this, I am confident that machine learning and deep learning will help investors make the right decisions when purchasing homes in various economic situations.

• Journal of Microbiology
• /
• v.45 no.6
• /
• pp.558-565
• /
• 2007

Using microarray analysis, we determined those Salmonella genes induced at the entry of stationary phase, and subsequently discovered that uncharacterized yliH was induced most dramatically. We set out to establish the molecular mechanism underlying the stationary phase induction of yliH under the standard culture condition, LB with vigorous aeration, by analyzing its promoter activity in various mutant backgrounds, lacking stationary phase ${\sigma}$, $RpoS^-$, or stringent signal molecules ppGpp, ${\Delta}relA$ ${\Delta}spoT$. It was found that the stationary phase induction of yliHp was partially dependent on rpoS but entirely dependent on ppGpp. DNA sequence analysis revealed that the Salmonella yliH gene is composed of 381 base-pair nucleotides, with overall amino acid sequence revealing 76.38% amino acid identity and 88.98% similarity with Escherichia coli yliH, although no motif from data base was noted for its possible role. Recently however, it has been reported that yliH in E. coli was implicated in biofilm formation and motility by repressing these activities (Domka et al., 2006). We have constructed a mutant Salmonella deleting yliH gene by allele replacement and examined its phenotype, and found that the yliH in Salmonella more or less affects motility and adherence by enhancing these activities. The effect on biofilm formation in Salmonella was uncertain. Moreover, addition of cloned yliH of E. coli into Salmonella did not reduce motility or adherence. Taken together, it appears that the pathways implicating yliH for biofilm formation and motility in E. coli and in Salmonella are somewhat different.

• Journal of Technology Innovation
• /
• v.14 no.1
• /
• pp.227-254
• /
• 2006

Under the recent trend of telecommunications and broadcasting convergence, much more various forms of new convergence broadcasting services are being introduced than ever before. Owing to the unique advantages, new convergence broadcasting services are expected to bring drastic changes to the current broadcasting service market. In this research, we attempt to examine what kind of attributes critically affect the competition among new convergence broadcasting services, and how much competitive they will be, based on the quantitative information about consumer preferences for the important attributes of new convergence broadcasting services. Conjoint survey was used in order to obtain stated preference data of consumers. From the results, some implications are obtained as follows. First, even though new convergence broadcasting services have many unique advantages, still price is the most important for the consumers. Second, it is expected that considerable consumer valuation exists for the unique advantages of new convergence broadcasting services like mobility and dual-way interactivity, which will add the competitiveness of those services in the future. Third, since midterm advertising puts negative utility on consumers, broadcasting services with midterm advertising will not be preferred to those with neither advertising nor midterm advertising. Fourth, service coverage and the number of consumers using the same broadcasting service have a significant feedback effect on the competition between broadcasting services from the dynamic aspect. Lastly, the consumer preference can be affected by demographic variables like age and gender, and broadcasting service usage patterns such as channel switchover for escaping advertisement and frequency of using other recorded media. Main findings of our research might become useful information for both telecommunication and broadcasting companies, contents providers, advertisers, and policy and regulation makers to cope with the uncertain environment of telecommunication and broadcasting convergence.