• ETRI Journal
• /
• 제43권5호
• /
• pp.775-786
• /
• 2021

To translate in real time, a simultaneous translation system should determine when to stop reading source tokens and generate target tokens corresponding to a partial source sentence read up to that point. However, conventional attention-based neural machine translation (NMT) models cannot produce translations with adequate latency in online scenarios because they wait until a source sentence is completed to compute alignment between the source and target tokens. To address this issue, we propose a reinforced learning (RL)-based attention mechanism, the reinforced attention mechanism, which allows a neural translation model to jointly train the stopping criterion and a partial translation model. The proposed attention mechanism comprises two modules, one to ensure translation quality and the other to address latency. Different from previous RL-based simultaneous translation systems, which learn the stopping criterion from a fixed NMT model, the modules can be trained jointly with a novel reward function. In our experiments, the proposed model has better translation quality and comparable latency compared to previous models.

• 스마트미디어저널
• /
• 제11권4호
• /
• pp.30-37
• /
• 2022

Disease threatens plant growth and recognizing the type of disease is essential to making a remedy. In recent years, deep learning has witnessed a significant improvement for this task, however, a large volume of labeled images is one of the requirements to get decent performance. But annotated images are difficult and expensive to obtain in the agricultural field. Therefore, designing an efficient and effective strategy is one of the challenges in this area with few labeled data. Transfer learning, assuming taking knowledge from a source domain to a target domain, is borrowed to address this issue and observed comparable results. However, current transfer learning strategies can be regarded as a supervised method as it hypothesizes that there are many labeled images in a source domain. In contrast, unsupervised transfer learning, using only images in a source domain, gives more convenience as collecting images is much easier than annotating. In this paper, we leverage unsupervised transfer learning to perform plant disease recognition, by which we achieve a better performance than supervised transfer learning in many cases. Besides, a vision transformer with a bigger model capacity than convolution is utilized to have a better-pretrained feature space. With the vision transformer-based unsupervised transfer learning, we achieve better results than current works in two datasets. Especially, we obtain 97.3% accuracy with only 30 training images for each class in the Plant Village dataset. We hope that our work can encourage the community to pay attention to vision transformer-based unsupervised transfer learning in the agricultural field when with few labeled images.

• International Journal of Computer Science & Network Security
• /
• 제23권11호
• /
• pp.99-109
• /
• 2023

The darknet is frequently referred to as the hub of illicit online activity. In order to keep track of real-time applications and activities taking place on Darknet, traffic on that network must be analysed. It is without a doubt important to recognise network traffic tied to an unused Internet address in order to spot and investigate malicious online activity. Any observed network traffic is the result of mis-configuration from faked source addresses and another methods that monitor the unused space address because there are no genuine devices or hosts in an unused address block. Digital systems can now detect and identify darknet activity on their own thanks to recent advances in artificial intelligence. In this paper, offer a generalised method for deep learning-based detection and classification of darknet traffic. Furthermore, analyse a cutting-edge complicated dataset that contains a lot of information about darknet traffic. Next, examine various feature selection strategies to choose a best attribute for detecting and classifying darknet traffic. For the purpose of identifying threats using network properties acquired from darknet traffic, devised a hybrid deep learning (DL) approach that combines Recurrent Neural Network (RNN) and Bidirectional LSTM (BiLSTM). This probing technique can tell malicious traffic from legitimate traffic. The results show that the suggested strategy works better than the existing ways by producing the highest level of accuracy for categorising darknet traffic using the Black widow optimization algorithm as a feature selection approach and RNN-BiLSTM as a recognition model.

• International journal of advanced smart convergence
• /
• 제13권2호
• /
• pp.48-60
• /
• 2024

With the exponential growth of satellite data utilization, machine learning has become pivotal in enhancing innovation and cybersecurity in satellite systems. This paper investigates the role of machine learning techniques in identifying and mitigating vulnerabilities and code smells within satellite software. We explore satellite system architecture and survey applications like vulnerability analysis, source code refactoring, and security flaw detection, emphasizing feature extraction methodologies such as Abstract Syntax Trees (AST) and Control Flow Graphs (CFG). We present practical examples of feature extraction and training models using machine learning techniques like Random Forests, Support Vector Machines, and Gradient Boosting. Additionally, we review open-access satellite datasets and address prevalent code smells through systematic refactoring solutions. By integrating continuous code review and refactoring into satellite software development, this research aims to improve maintainability, scalability, and cybersecurity, providing novel insights for the advancement of satellite software development and security. The value of this paper lies in its focus on addressing the identification of vulnerabilities and resolution of code smells in satellite software. In terms of the authors' contributions, we detail methods for applying machine learning to identify potential vulnerabilities and code smells in satellite software. Furthermore, the study presents techniques for feature extraction and model training, utilizing Abstract Syntax Trees (AST) and Control Flow Graphs (CFG) to extract relevant features for machine learning training. Regarding the results, we discuss the analysis of vulnerabilities, the identification of code smells, maintenance, and security enhancement through practical examples. This underscores the significant improvement in the maintainability and scalability of satellite software through continuous code review and refactoring.

• 아태비즈니스연구
• /
• 제12권4호
• /
• pp.13-40
• /
• 2021

Purpose - The purpose of this study is to predict corporate corruption in emerging markets such as Brazil, Russia, India, and China (BRIC) using different machine learning techniques. Since corruption is a significant problem that can affect corporate performance, particularly in emerging markets, it is important to correctly identify whether a company engages in corrupt practices. Design/methodology/approach - In order to address the research question, we employ predictive analytic techniques (machine learning methods). Using the World Bank Enterprise Survey Data, this study evaluates various predictive models generated by seven supervised learning algorithms: k-Nearest Neighbour (k-NN), Naïve Bayes (NB), Decision Tree (DT), Decision Rules (DR), Logistic Regression (LR), Support Vector Machines (SVM), and Artificial Neural Network (ANN). Findings - We find that DT, DR, SVM and ANN create highly accurate models (over 90% of accuracy). Among various factors, firm age is the most significant, while several other determinants such as source of working capital, top manager experience, and the number of permanent full-time employees also contribute to company corruption. Research implications or Originality - This research successfully demonstrates how machine learning can be applied to predict corporate corruption and also identifies the major causes of corporate corruption.

• 반도체디스플레이기술학회지
• /
• 제22권1호
• /
• pp.113-117
• /
• 2023

As sputter equipment becomes more complex, it becomes increasingly difficult to understand the parameters that affect the thickness uniformity of thin metal film deposited by sputter. To address this issue, we verified a deep learning model that can predict complex relationships. Specifically, we trained the model to predict the height of 36 magnets based on the thickness of the material, using Support Vector Machine (SVM), Multilayer Perceptron (MLP), 1D-Convolutional Neural Network (1D-CNN), and 2D-Convolutional Neural Network (2D-CNN) algorithms. After evaluating each model, we found that the MLP model exhibited the best performance, especially when the dataset was constructed regardless of the thin film material. In conclusion, our study suggests that it is possible to predict the sputter equipment source using film thickness data through a deep learning model, which makes it easier to understand the relationship between film thickness and sputter equipment.

• 데이타베이스연구회지:데이타베이스연구
• /
• 제34권3호
• /
• pp.137-147
• /
• 2018

기계 학습을 통한 인간 동작 인지 (human activity recognition) 시스템에서 중요한 요소는 충분한 양의 라벨 데이터 (labeled data)를 확보하는 것이다. 그러나 라벨 데이터를 확보하는 일은 많은 비용과 시간을 필요로 한다. 매우 적은 수의 라벨 데이터를 가지고 있는 새로운 환경 (타겟 도메인)에서 동작 인지 시스템을 구축하는 경우, 기존의 환경 (소스 도메인)의 데이터나 이 환경에서 학습된 분류기(classifier)를 사용하는 것은 도메인이 서로 다르기 때문에 바람직하지 않다. 기존의 기계 학습 방법들이 이러한 문제를 해결할 수 없으므로 전이 학습 (transfer learning) 방법이 제시되었으며, 이 방법에서는 소스 도메인에서 확보한 지식을 활용하여 타겟 도메인에서의 분류기 성능을 높이도록 하고 있다. 본 논문에서는 다중 태스크 신경망 (multitask neural network)을 사용하여 매우 제한된 수의 데이터만으로 정확도가 높은 동작 인지 분류기를 생성하는 전이 학습방법을 제안한다. 이 방법에서는 소스 및 타겟 도메인 분류기의 손실 함수 최소화가 별개의 태스크로 간주된다. 즉, 하나의 신경망을 사용하여 두 태스크의 손실 함수를 동시에 최소화하는 방식으로 지식 전이(knowledge transfer)가 일어나게 된다. 또한, 제안한 방법에서는 모델 학습을 위하여 비지도 방식(unsupervised manner)으로 라벨이 부여되지 않은 데이터를 활용한다. 실험 결과, 제안한 방법은 기존의 방법에 비하여 일관적으로 우수한 성능을 보여주고 있다.

• 한국산학기술학회논문지
• /
• 제18권11호
• /
• pp.46-52
• /
• 2017

최근 ICT 및 IoT 제품의 활용 분야가 다양화 되면서 오픈소스 소프트웨어의 활용 분야가 컴퓨터, 스마트폰, IoT 디바이스 등 다양한 기기와 환경에서 활용되고 있다. 이처럼 오픈소스 소프트웨어의 활용분야가 다양해짐에 따라 오픈소스의 보안 취약점을 악용하는 불법적인 사례가 지속적으로 증가하고 있다. 이에 따라 다양한 시큐어 코딩을 위한 도구나 프로그램이 출시되고 활용되고 있지만 여전히 많은 취약점들이 발생하고 있다. 본 논문에서는 안전한 오픈 소스 소프트웨어 개발을 위해 오픈 소스의 취약점 분석 결과에 의한 이력과 패턴을 지속적으로 학습하여 신규 취약점 분석에 활용할 수 있는 방법을 제안한다. 본 연구를통해 취약점 이력 및 패턴 학습기반의 취약점 분석 시스템을 설계하였으며, 프로토타입으로 구현하여 실험을 통해 시스템의 성능을 평가하였다. 5개의 취약점 항목에 대해 평균 취약점 검출 시간은 최대 약 1.61sec가 단축되었으며, 평균 검출 정확도는 약 44%point가 향상된 것을 평가결과에서 확인할 수 있었다. 본 논문의 내용 및 결과는 소프트웨어 취약점 연구 분야에 대한 발전과 소프트웨어 개발자들의 취약점 분석을 통한 시큐어 코딩에 도움이 될 것을 기대한다.

• 한국컴퓨터정보학회지
• /
• 제14권2호
• /
• pp.235-244
• /
• 2006

본 논문은 VoIP 서비스 네트워크에서 UDP, ICMP, Echo, TCP Syn 패킷 등을 조작한 SYN Flooding 방법, TCP/UDP을 이용한 어플리케이션을 통한 DoS 공격, IP Source Address Spoofing과 Smurf을 이용한 공격, 웜과 트로이목마를 혼합한 알려지지 않는 DoS 공격을 하였다. IPS에서 방어를 위한 동적 업데이트 엔진의 필요성을 정의하고, 엔진의 설계 시에 내 외부의 RT통계에서 트래픽 양을 측정하며, 학습모듈과 통계적 공격에 대한 퍼지 로직 엔진모듈을 설계한다. 엔진은 3가지 공격 등급(Attack, Suspicious, Normal)을 판단하여, Footprint Lookup 모듈에서 AND나 OR 연산을 통해 최적의 필터링 엔진 상태를 유지한다. 실험을 통해 IPS 차단 모듈과 필터링엔진의 실시간 업데이트되어 DoS 공격의 차단이 수행됨을 확인하였다. 실시간 동적으로 업데이트되는 엔진과 필터는 DoS 공격으로부터 VOIP 서비스를 보호하여 유비쿼터스 보안성을 강화시킨 것으로 판명되어졌다.

• 한국컴퓨터정보학회논문지
• /
• 제11권6호
• /
• pp.165-174
• /
• 2006

본 논문은 VoIP 서비스 네트워크에서 UDP, ICMP, Echo, TCP Syn 패킷 등을 조작한 SYN Flooding 방법, TCP/UDP을 이용한 어플리케이션을 통한 DoS 공격, IP Source Address Spoofing과 Smurf을 이용한 공격, 웜과 트로이목마를 혼합한 알려지지 않는 DoS 공격을 하였다. IPS에서 방어를 위한 동적 업데이트 엔진의 필요성을 정의하고, 엔진의 설계 시에 내 외부의 RT통계에서 트래픽 양을 측정하며, 학습모듈과 통계적 공격에 대한 퍼지 로직 엔진모듈을 설계한다. 엔진은 3가지 공격 등급(Attack, Suspicious, Normal)을 판단하여, Footprint Lookup 모듈에서 AND나 OR 연산을 통해 최적의 필터링 엔진 상태를 유지한다. 실험을 통해 IPS 차단 모듈과 필터링엔진의 실시간 업데이트되어 DoS 공격의 차단이 수행됨을 확인하였다. 실시간 동적으로 업데이트되는 엔진과 필터는 DoS 공격으로부터 VoIP 서비스를 보호하여 유비쿼터스 보안성을 강화시킨 것으로 판명되어졌다.