• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.32 no.2
• /
• pp.267-277
• /
• 2022

Due to the increasing proportion of cloud and remote working environments, various information security incidents are occurring. Insider threats have emerged as a major issue, with cases in which corporate insiders attempting to leak confidential data by accessing it remotely. In response, insider threat detection approaches based on machine learning have been developed. However, existing machine learning methods used to detect insider threats do not take biases and variances into account, which leads to limited performance. In this paper, boosting-type ensemble learning algorithms are applied to verify the performance of malicious insider detection, conduct a close analysis, and even consider the imbalance in datasets to determine the final result. Through experiments, we show that using ensemble learning achieves similar or higher accuracy to other existing malicious insider detection approaches while considering bias-variance tradeoff. The experimental results show that ensemble learning using bagging and boosting methods reached an accuracy of over 98%, which improves malicious insider detection performance by 5.62% compared to the average accuracy of single learning models used.

• The Journal of the Convergence on Culture Technology
• /
• v.8 no.5
• /
• pp.613-618
• /
• 2022

Research on artificial intelligence based on SNN (Spiking Neural Networks) is drawing attention as a next-generation artificial intelligence that can overcome the limitations of artificial intelligence based on DNN (Deep Neural Networks) that is currently popular. In this paper, we describe the structure of the SNN compiler, a system SW that generate code from SNN description for neuromorphic computing systems. We also introduce the algorithms used for compiler implementation and present experimental results on how the execution time varies in neuromorphic computing systems depending on the the mapping algorithm. The mapping algorithm proposed in the text showed a performance improvement of up to 3.96 times over a random mapping. The results of this study will allow SNNs to be applied in various neuromorphic hardware.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.05a
• /
• pp.441-443
• /
• 2022

Currently, image processing is being used in various fields such as image analysis, image recognition, and factory automation according to the development of IT technology. Salt and pepper noise is generated due to various external factors in the process of acquiring or transmitting an image, which deteriorates the image quality. Therefore, noise removal to improve image quality is essential. Various methods have been proposed to remove salt and pepper noise, and representative examples include AF, MF, and A-TMF. However, the conventional filter has insufficient noise removal performance in a high-density noise environment. Therefore, in this paper, we propose an algorithm for estimating and processing the original pixel by using the modified distance weight filter only in the case of noise, and replacing the original pixel in case of non-noise after performing noise judgment. To evaluate the performance of the proposed algorithm, we compare and analyze it with existing algorithms using PSNR.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.10a
• /
• pp.185-187
• /
• 2022

Because of the rapid change in the educational paradigm in the Fourth Industrial Revolution Era, Artificial Intelligence (AI) Education is becoming increasingly important today. The 2022 Revised Curriculum focuses on AI Education that can cultivate the fundamental skills and competencies needed in the future society. The following are the directions presented in this study for improving computational thinking and mathematical thinking in AI Education in elementary and secondary schools. First, studying teaching principles that allow students to understand AI concepts and principles and develop their ability to solve real-life problems is necessary in terms of computational thinking skills education. Second, an educational program is required for students to acquire algorithms using formulas and learn principles in the process of computers thinking like humans as part of their mathematical thinking ability to understand AI. A study on expectations through the analysis of competent learning effects that may arise from the relationship between instructors and learners was proposed as a future research project.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.33 no.5
• /
• pp.709-719
• /
• 2023

Recently, Post-Quantum Cryptography (PQC), which is secure against attacks using quantum computers, has been actively studied. In 2022, the KpqC competition, a competition for domestic PQC standardization, was launched, and a total of 16 candidate algorithms were received, and the first round is underway. In this paper, we apply Alexander May's Meet-LWE attack to TiGER, a lattice-based key encapsulation mechanism that is a candidate for the first round of the KpqC competition, and analyze its concrete attack complexity. The computational results of applying the Meet-LWE attack to each of the proposed parameters of TiGER show that the proposed TiGER192 parameter, which targets 192-bit quantum security, actually achieves 170-bit classical security. In addition, we propose a parameter setting to increase the attack complexity against the Meet-LWE attack.

• The Journal of the Convergence on Culture Technology
• /
• v.9 no.6
• /
• pp.131-136
• /
• 2023

Local youth outmigration is increasingly growing. Various studies are being conducted to identify the factors contributing to this problem, but there is a lack of research analyzing each region individually. Therefore, this study aims to analyze the factors influencing youth outmigration in Busan and predict the risk levels of youth population outflow using machine learning techniques. By utilizing district-level data collected from the KOSIS, we divided the population into three groups based on age (the early 20s, late 20s, and early 30s) and employed Decision Tree and Random Forest algorithms to classify and predict the risk levels of youth population outmigration. The results indicate that the predictive model for youth outmigration risk levels achieves the highest accuracies of 0.93, 0.75, and 0.63 for each age group, respectively.

• Journal of Digital Policy
• /
• v.3 no.1
• /
• pp.27-32
• /
• 2024

In this paper, proposes a super resolution method that enhances the quality of results by refining texture features, contrasting each, and utilizing the results as weights. For the improvement of quality, a precise and clear restoration result in details such as boundary areas is crucial in super resolution, along with minimizing unnecessary artifacts like noise. The proposed method constructs a residual block structure with multiple paths and skip-connections for feature estimation in conventional Convolutional Neural Network (CNN)-based super resolution methods to enhance quality. Additional learning is performed for sharpened and blurred image results for further texture analysis. By contrasting each super resolution result and allocating weights through this process, the proposed method achieves improved quality in detailed and smoothed areas of the image. The experimental results of the proposed method, evaluated using the PSNR and SSIM values as quality metrics, show higher results compared to existing algorithms, confirming the enhancement in quality.

• Tunnel and Underground Space
• /
• v.34 no.2
• /
• pp.143-153
• /
• 2024

Tunnel Boring Machine (TBM) method is a tunnel excavation method that produces lower levels of noise and vibration during excavation compared to drilling and blasting methods, and it offers higher stability. It is increasingly being applied to tunnel projects worldwide. The disc cutter is an excavation tool mounted on the cutterhead of a TBM, which constantly interacts with the ground at the tunnel face, inevitably leading to wear. In this study quantitatively predicted disc cutter wear using geological conditions, TBM operational parameters, and machine learning algorithms. Among the input variables for predicting disc cutter wear, the Uniaxial Compressive Strength (UCS) is considerably limited compared to machine and wear data, so the UCS estimation for the entire section was first conducted using TBM machine data, and then the prediction of the Coefficient of Wearing rate(CW) was performed with the completed data. Comparing the performance of CW prediction models, the XGBoost model showed the highest performance, and SHapley Additive exPlanation (SHAP) analysis was conducted to interpret the complex prediction model.

• Proceedings of the KSR Conference
• /
• 2011.10a
• /
• pp.2409-2421
• /
• 2011

This paper will present hybrid tests to a one bay-one story steel frame structure under ground excitation. A structure used in this paper for hybrid test, to evaluate performance and behavior, is divided into two models; one is numerical model with one column element, and a truss or a beam element, the other is physical substructural model with one beam-column element. All tests considered one or three degrees of freedom to implement real-time hybrid test, and two control algorithms to control hardware are used; one using MATLAB/Simulink, the other using OpenSees, OpenFresco and xPCTarget. In addition, for real-time data communication between numerical and physical substructural models SCRAMNet was used. The results of hybrid tests were compared with one of numerical analysis of numerical model with fiber force-based beam-column elements using OpenSees. Real-time hybrid tests were implemented for the validation of control system with simple structure, and then it will be extended to hybrid test for higher nonlinear or complex structure later on.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.12 no.6
• /
• pp.531-536
• /
• 2002

In this paper, we have applied Genetic Algorithms(GAs) to Intrusion Detection System(TDS), and then proposed and simulated the misuse detection model firstly. We have implemented with the KBD contest data, and tried to simulated in the same environment. In the experiment, the set of record is regarded as a chromosome, and GAs are used to produce the intrusion patterns. That is, the intrusion rules are generated. We have concentrated on the simulation and analysis of classification among the Data Mining techniques and then the intrusion patterns are produced. The generated rules are represented by intrusion data and classified between abnormal and normal users. The different rules are generated separately from three models "Time Based Traffic Model", "Host Based Traffic Model", and "Content Model". The proposed system has generated the update and adaptive rules automatically and continuously on the misuse detection method which is difficult to update the rule generation. The generated rules are experimented on 430M test data and almost 94.3% of detection rate is shown.3% of detection rate is shown.