• Journal of Information Technology Services
• /
• v.18 no.1
• /
• pp.173-186
• /
• 2019

Recently, as the amount of fine dust has risen rapidly, our interest is increasing day by day. It is virtually impossible to remove fine dust. However, it is best to predict the concentration of fine dust and minimize exposure to it. In this study, we developed a mathematical model that can predict the concentration of fine dust using various information related to the weather and air quality, which is provided in real time in 'Air Korea (http://www.airkorea.or.kr/)' and 'Weather Data Open Portal (https://data.kma.go.kr/).' In the mathematical model, various domestic seasonal variables and atmospheric state variables are extracted by multiple regression analysis. The parameters that have significant influence on the fine dust concentration are extracted, and using ANN (Artificial Neural Network) and SVM (Support Vector Machine), which are machine learning techniques, we proposed a prediction model. The proposed model can verify its effectiveness by using past dust and weather big data.

• Journal of Information Processing Systems
• /
• v.15 no.5
• /
• pp.1231-1242
• /
• 2019

This paper presents a two-dimensional attention-based long short-memory (2D-ALSTM) model for stock index prediction, incorporating input attention and temporal attention mechanisms for weighting of important stocks and important time steps, respectively. The proposed model is designed to overcome the long-term dependency, stock selection, and stock volatility delay problems that negatively affect existing models. The 2D-ALSTM model is validated in a comparative experiment involving the two attention-based models multi-input LSTM (MI-LSTM) and dual-stage attention-based recurrent neural network (DARNN), with real stock data being used for training and evaluation. The model achieves superior performance compared to MI-LSTM and DARNN for stock index prediction on a KOSPI100 dataset.

• Journal of Information Processing Systems
• /
• v.15 no.5
• /
• pp.1029-1035
• /
• 2019

In recent years, artificial intelligence (AI) services have become one of the most essential parts to extend human capabilities in various fields such as face recognition for security, weather prediction, and so on. Various learning algorithms for existing AI services are utilized, such as classification, regression, and deep learning, to increase accuracy and efficiency for humans. Nonetheless, these services face many challenges such as fake news spread on social media, stock selection, and volatility delay in stock prediction systems and inaccurate movie-based recommendation systems. In this paper, various algorithms are presented to mitigate these issues in different systems and services. Convolutional neural network algorithms are used for detecting fake news in Korean language with a Word-Embedded model. It is based on k-clique and data mining and increased accuracy in personalized recommendation-based services stock selection and volatility delay in stock prediction. Other algorithms like multi-level fusion processing address problems of lack of real-time database.

• Annual Conference of KIPS
• /
• 2013.11a
• /
• pp.977-979
• /
• 2013

본 논문은 스마트 폰 상의 SNS(Social Network Service)기반의 집단의 지식을 효율적으로 공유할 수 있도록 게임화(Gamification)전략을 도입한 질의응답 시스템을 제안한다. 질의응답 시스템은 신규 사용자 및 이미 많은 사용자가 있는 트위터와 페이스북 서비스에서 제공하는 API를 이용하여 사용자를 확보하고, 질문과 답변을 쉽게 할 수 있도록 한다. 또한 게이미피케이션 전략을 통해 사용자의 적절한 서비스 재 몰입루프를 형성한다. 따라서 제안된 시스템에 따라 서비스의 장기간 사용을 사용자에게 효율적으로 유도할 수 있다.

• Annual Conference of KIPS
• /
• 2013.11a
• /
• pp.453-456
• /
• 2013

최근 스마트폰 보급률 증가 및 무선 센서 네트워크(Wireless Sensor Networks) 기술 발전에 따라 해당 기술을 화재감시, 침입 감지와 같은 응용에 융합하는 연구가 활발히 진행되고 있다. 하지만 기존 연구들은 주기를 기반으로 감지를 수행하기 때문에 화재 및 침입 판단이 지연되는 문제점이 존재한다. 이를 위해, 본 논문에서는 판단 주기를 동적으로 설정하는 조기 화재 판단 알고리즘을 통해 화재 판단 시간을 단축시켜 빠른 대처를 할 수 있도록 지원하는 새로운 화재감시 및 침입 감지 시스템을 개발한다. 아울러, 적외선 센서를 이용하여 무단 침입을 감지함으로써 도난 및 파손과 함께 방화로 인한 화재를 방지할 수 있다. 마지막으로 성능평가를 통해 제안하는 시스템이 화재 판단 측면에서 기존 연구보다 우수함을 입증한다.

• Annual Conference of KIPS
• /
• 2013.11a
• /
• pp.702-704
• /
• 2013

네트워크의 급속한 발전으로 데이터가 방대해짐에 따라 정보시스템의 역기능으로 네트워크 공격이 다양해지고 빈번하게 발생하면서 네트워크 침입 탐지시스템 기능과 패킷 차단 기능이 중요시되고 있다. 본 논문에서는 침입 탐지 시스템 오픈 소스인 Snort와 리눅스의 iptables 시스템의 각 장점을 활용하여 연동하는 snort-inline과는 다른 방식을 사용하여 실시간으로 침입 탐지시스템의 역할을 하는 스크립트를 Python으로 구현하였다. 구현한 시스템의 성능 검증을 위해 공격자가 해킹 대상 시스템에 DOS 공격을 하여 구현 모듈에서 snort의 탐지 능력과 iptables의 패킷 차단 명령문이 실행되어 악의적 패킷 접근을 차단할 수 있음을 제시하였다.

• Journal of Positioning, Navigation, and Timing
• /
• v.10 no.3
• /
• pp.159-168
• /
• 2021

Precise Point Positioning-Real Time Kinematic (PPP-RTK) is an improved PPP method that provides the user receiver with satellite code and phase bias correction information in addition to the satellite orbit and clock, thus enabling single-receiver ambiguity resolution. Single station PPP-RTK concept is special case of PPP-RTK in that corrections are computed, instead of a network, by only one single GNSS receiver. This study is performed to experimentally verify the positioning accuracy performance of single baseline RTK level by a user who utilizes correction for a single station PPP-RTK using dual frequencies. As an experimental result, the horizontal and vertical 95% accuracy was 2.2 cm, 4.4 cm, respectively, which verify the same performance as the single baseline RTK.

• Nuclear Engineering and Technology
• /
• v.54 no.4
• /
• pp.1230-1244
• /
• 2022

This paper proposes an algorithm for signal validation using unsupervised methods in emergency situations at nuclear power plants (NPPs) when signals are rapidly changing. The algorithm aims to determine the stuck failures of signals in real time based on a variational auto-encoder (VAE), which employs unsupervised learning, and long short-term memory (LSTM). The application of unsupervised learning enables the algorithm to detect a wide range of stuck failures, even those that are not trained. First, this paper discusses the potential failure modes of signals in NPPs and reviews previous studies conducted on signal validation. Then, an algorithm for detecting signal failures is proposed by applying LSTM and VAE. To overcome the typical problems of unsupervised learning processes, such as trainability and performance issues, several optimizations are carried out to select the inputs, determine the hyper-parameters of the network, and establish the thresholds to identify signal failures. Finally, the proposed algorithm is validated and demonstrated using a compact nuclear simulator.

• Journal of Auto-vehicle Safety Association
• /
• v.14 no.3
• /
• pp.71-76
• /
• 2022

In autonomous driving cars, 3D pose estimation can be one of the effective methods to enhance safety control for OOP (Out of Position) passengers. There have been many studies on human pose estimation using a camera. Previous methods, however, have limitations in automotive applications. Due to unexplainable failures, CNN methods are unreliable, and other methods perform poorly. This paper proposes robust real-time multi-human 3D pose estimation architecture in vehicle using monocular RGB camera. Using particle filter, our approach integrates CNN 2D/3D pose measurements with available information in vehicle. Computer simulations were performed to confirm the accuracy and robustness of the proposed algorithm.

• Nuclear Engineering and Technology
• /
• v.54 no.5
• /
• pp.1747-1753
• /
• 2022

Compton imaging is the main method for locating radioactive hot spots emitting high-energy gamma-ray photons. In particular, this imaging method is crucial when the photon energy is too high for coded-mask aperture imaging methods to be effective or when a large field of view is required. Reconstruction of the photon source requires advanced Compton event processing algorithms to determine the exact position of the source. In this study, we introduce a novel method based on a Deep Learning algorithm with a Convolutional Neural Network (CNN) to perform Compton imaging. This algorithm is trained on simulated data and tested on real data acquired with Caliste, a single planar CdTe pixelated detector. We show that performance in terms of source location accuracy is equivalent to state-of-the-art algorithms, while computation time is significantly reduced and sensitivity is improved by a factor of ~5 in the Caliste configuration.