• Annual Conference of KIPS
• /
• 2021.05a
• /
• pp.382-384
• /
• 2021

AI (Artificial Intelligence)기술이 상용화되면서 최근 기업들은 AI 모델의 기능을 서비스화하여 제공하고 있다. 하지만 최근 이러한 서비스를 이용하여 기업이 자본을 투자해 학습시킨 AI 모델을 탈취하는 공격이 등장하여 위협이 되고 있다. 본 논문은 최근 연구되고 있는 이러한 모델 탈취 공격들에 대해 공격자의 정보를 기준으로 분류하여 서술한다. 또한 본 논문에서는 모델 탈취 공격에 대응하기 위해 다양한 관점에서 시도되는 방어 기법들에 대해 서술한다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2023.07a
• /
• pp.715-718
• /
• 2023

본 연구는 ChatGPT 모델의 특성과 장점을 활용하여 프롬프트 엔지니어링 기법을 연구하고자 하였다. 프롬프트는 엔지니어가 원하는 결과를 잘 얻을 수 있도록 하는 것이 목표이기 때문에 ChatGPT와 프롬프트 엔지니어링의 상호작용과 효과적인 프롬프트 엔지니어링 기법을 개발할 필요가 있다. 연구 방법으로는 ChatGPT에 대한 학습자 사전 설문조사에서 학습자를 분석하였고, 이를 반영하여 프로그래밍 문제를 제시하고 해결하는 과정을 거치면서 다양한 ChatGPT 사용에 대한 분석과 학습자 분석이 이루어졌다. 그 결과 비전공자가 듣고 있는 프로그래밍 수업에서 ChatGPT를 활용하여 얻은 통찰력으로 프롬프트에 필요한 가이드 라인을 마련하였다. 본 연구를 기반으로 향후 비전공자를 위한 파이썬 프로그래밍 수업에서 ChatGPT를 활용한 수업모델을 제시하고 학습자의 피드백 또는 적응형 학습에 활용할 수 있는 방법을 모색할 것이다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2023.07a
• /
• pp.51-54
• /
• 2023

최근 딥러닝 기술의 발달로 물체 탐지를 위한 객체 인식 분야가 기계학습을 접목한 연구가 급격히 증가하고 있다. 하지만, 탐지하려는 물체가 다른 객체에 가려진 경우와 같이 특수한 상황에 대한 데이터의 수량이 부족하여 성능 저하를 야기한다는 점과, 객체 탐지 수행 과정에서 작은 객체의 탐지가 어렵다는 한계점이 있다. 본 연구는 전술한 문제점을 보완할 방법을 제안한다. 데이터 증강 기법을 이용하여 클래스가 부족한 데이터의 양을 늘려 학습 데이터를 증강시켰다. 한편, SRGAN을 사용하여 작은 객체를 확대시킨 뒤 이미지를 합성시켜 데이터를 구성하였다. 제안된 방법은 PyTorch 환경에서 YOLOv5를 수행한 결과, 객체 탐지 성능이 향상되는 것을 확인할 수 있었다.

• Journal of Ocean Engineering and Technology
• /
• v.37 no.5
• /
• pp.181-189
• /
• 2023

The fuel consumption of marine diesel engines holds paramount importance in contemporary maritime transportation and shapes energy efficiency strategies of ocean-going vessels. Nonetheless, a noticeable gap in knowledge prevails concerning the influence of ship hull conditions and propeller roughness on fuel consumption. This study bridges this gap by utilizing artificial intelligence techniques in Matlab, particularly convolutional neural networks (CNNs) to comprehensively investigate these factors. We propose a time-series prediction model that was built on numerical simulations and aimed at forecasting ship hull and propeller conditions. The model's accuracy was validated through a meticulous comparison of predictions with actual ship-hull and propeller conditions. Furthermore, we executed a comparative analysis juxtaposing predictive outcomes with navigational environmental factors encompassing wind speed, wave height, and ship loading conditions by the fuzzy clustering method. This research's significance lies in its pivotal role as a foundation for fostering a more intricate understanding of energy consumption within the realm of maritime transport.

• International journal of advanced smart convergence
• /
• v.12 no.4
• /
• pp.111-118
• /
• 2023

In the ever-changing landscape of finance, the fusion of artificial intelligence (AI)and pair trading strategies has captured the interest of investors and institutions alike. In the context of supervised machine learning, crafting precise and accurate labels is crucial, as it remains a top priority to empower AI models to surpass traditional pair trading methods. However, prevailing labeling techniques in the financial sector predominantly concentrate on individual assets, posing a challenge in aligning with pair trading strategies. To address this issue, we propose an inventive approach that melds the Triple Barrier Labeling technique with pair trading, optimizing the resultant labels through genetic algorithms. Rigorous backtesting on cryptocurrency datasets illustrates that our proposed labeling method excels over traditional pair trading methods and corresponding buy-and-hold strategies in both profitability and risk control. This pioneering method offers a novel perspective on trading strategies and risk management within the financial domain, laying a robust groundwork for further enhancing the precision and reliability of pair trading strategies utilizing AI models.

• Journal of Drive and Control
• /
• v.20 no.4
• /
• pp.54-63
• /
• 2023

High and intermittent loading cycles induce fatigue damage to transmission components, resulting in premature gearbox failure. To identify gearbox defects, numerous vibration-based diagnostics techniques, using several artificial intelligence (AI) algorithms, have recently been presented. In this paper, an optimized deep belief network (DBN) model for gearbox problem diagnosis was designed based on time-frequency visual pattern identification. To optimize the hyperparameters of the model, a particle swarm optimization (PSO) approach was integrated into the DBN. The proposed model was tested on two gearbox datasets: a wind turbine gearbox and an experimental gearbox. The optimized DBN model demonstrated strong and robust performance in classification accuracy. In addition, the accuracy of the generated datasets was compared using traditional ML and DL algorithms. Furthermore, the proposed model was evaluated on different partitions of the dataset. The results showed that, even with a small amount of sample data, the optimized DBN model achieved high accuracy in diagnosis.

• ETRI Journal
• /
• v.45 no.6
• /
• pp.929-938
• /
• 2023

The detection of all the symbols transmitted simultaneously in multiuser systems using limited wireless resources is challenging. Traditional model-based methods show high performance with perfect channel state information (CSI); however, severe performance degradation will occur if perfect CSI cannot be acquired. In contrast, data-driven methods perform slightly worse than model-based methods in terms of symbol error ratio performance in perfect CSI states; however, they are also able to overcome extreme performance degradation in imperfect CSI states. This study proposes a novel deep learning-based method by improving a state-of-the-art data-driven technique called deep soft interference cancellation (DSIC). The enhanced DSIC (EDSIC) method detects multiuser symbols in a fully sequential manner and uses an efficient neural network structure to ensure high performance. Additionally, error-propagation mitigation techniques are used to ensure robustness against channel uncertainty. The EDSIC guarantees a performance that is very close to the optimal performance of the existing model-based methods in perfect CSI environments and the best performance in imperfect CSI environments.

• ETRI Journal
• /
• v.46 no.2
• /
• pp.263-276
• /
• 2024

Different diseases occur in the brain. For instance, hereditary and progressive diseases affect and degenerate the white matter. Although addressing, diagnosing, and treating complex abnormalities in the brain is challenging, different strategies have been presented with significant advances in medical research. With state-of-art developments in artificial intelligence, new techniques are being applied to brain magnetic resonance images. Deep learning has been recently used for the segmentation and classification of brain images. In this study, we classified normal and pathological brain images using pretrained deep models through transfer learning. The EfficientNet-B5 model reached the highest accuracy of 98.39% on real data, 91.96% on augmented data, and 100% on pathological data. To verify the reliability of the model, fivefold cross-validation and a two-tier cross-test were applied. The results suggest that the proposed method performs reasonably on the classification of brain magnetic resonance images.

• Korean Journal of Radiology
• /
• v.21 no.4
• /
• pp.387-401
• /
• 2020

Radiomics and deep learning have recently gained attention in the imaging assessment of various liver diseases. Recent research has demonstrated the potential utility of radiomics and deep learning in staging liver fibroses, detecting portal hypertension, characterizing focal hepatic lesions, prognosticating malignant hepatic tumors, and segmenting the liver and liver tumors. In this review, we outline the basic technical aspects of radiomics and deep learning and summarize recent investigations of the application of these techniques in liver disease.

• International journal of advanced smart convergence
• /
• v.13 no.2
• /
• pp.88-93
• /
• 2024

The main problem with existing edge detection techniques is that they have many limitations in detecting edges for complex and diverse images that exist in the real world. This is because only edges of a defined shape are discovered based on an accurate definition of the edge. One of the methods to solve this problem is the cost minimization method. In the cost minimization method, cost elements and cost functions are defined and used. The cost function calculates the cost for the candidate edge model generated according to the candidate edge generation strategy, and if the cost is found to be satisfactory, the candidate edge model becomes the edge for the image. In this study, we proposed an enhanced candidate edge generation strategy to discover edges for more diverse types of images in order to improve the shortcoming of the cost minimization method, which is that it only discovers edges of a defined type. As a result, improved edge detection results were confirmed.