• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.10 no.4
• /
• pp.11-23
• /
• 2011

The objective of this paper is to develop a daily pattern clustering algorithm using historical traffic data that can reliably detect under various traffic flow conditions in urban streets. The developed algorithm in this paper is categorized into two major parts, that is to say a macroscopic and a microscopic points of view. First of all, a macroscopic analysis process deduces a daily peak/non-peak hour and emphasis analysis time zones based on the speed time-series. A microscopic analysis process clusters a daily pattern compared with a similarity between individuals or between individual and group. The name of the developed algorithm in microscopic analysis process is called "Two-step speed clustering (TSC) algorithm". TSC algorithm improves the accuracy of a daily pattern clustering based on the time-series speed variation data. The experiments of the algorithm have been conducted with point detector data, installed at a Ansan city, and verified through comparison with a clustering techniques using SPSS. Our efforts in this study are expected to contribute to developing pattern-based information processing, operations management of daily recurrent congestion, improvement of daily signal optimization based on TOD plans.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.13 no.2
• /
• pp.1-11
• /
• 2014

In this paper, travel speed patterns were deducted based on historical DSRC travel speed data using Decision Tree technique to improve availability of the massive amount of historical data. These patterns were designed to reflect spatio-temporal vicissitudes in reality by generating pattern units classified by months, time of day, and highway sections. The study area was from Seoul TG to Ansung IC sections on Gyung-bu highway where high peak time of day frequently occurs in South Korea. Decision Tree technique was applied to categorize travel speed according to day of week. As a result, five different pattern groups were generated: (Mon)(Tue Wed Thu)(Fri)(Sat)(Sun). Statistical verification was conducted to prove the validity of patterns on nine different highway sections, and the accuracy of fitting was found to be 93%. To reduce travel pattern errors against individual travel speed data, inclusion of four additional variables were also tested. Among those variables, 'traffic condition on previous month' variable improved the pattern grouping accuracy by reducing 50% of speed variance in the decision tree model developed.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.40 no.3
• /
• pp.575-585
• /
• 2015

The traffic classification is a preliminary and essential step for stable network service provision and efficient network resource management. However, the payload signature-based method has significant drawbacks in high-speed network environment that the processing speed is much slower than other methods such as header-based and statistical methods. In addition, as signature numbers are increasing, traffic analysis speed also declines because of signature matching method that does not consider analytic efficiency of each signature and traffic occurrence feature. In this paper, we propose a signature list reordering method in order by analytic value of each signature. When we reordered the signature list by the proposed method, we achieved about 30% improvement in speed of the traffic analysis compared with random signature list.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.15 no.1
• /
• pp.252-261
• /
• 1991

A particle trajectory model to simulate two-phase particle-laden crossjets into two-dimensional horizontal free stream has been developed to study the variations of the jet trajectories and velocity variations of the gaseous and the particulate phases. The following conclusions may be drawn from the predicted results, which are in agreement with experimental observations. The penetration of the two-phase jet in a crossflow is greater than that of the single-phase jet. The penetration of particles into the free stream increases with increasing particle size, solids-gas loading ratio and carrier gas to free stream velocity ratio at the jet exit. When the particle size is large, the solid particles separate from the carrier gas , while the particles are completely suspended in the carrier gas for the case of small size particles. As the particle to carrier gas velocity ratio at the jet exit is less than unity, the particles in the vicinity of the jet exit are accelerated by the carrier gas. As the injection angle is increased, the difference of the particle trajectory from that of the pure gas becomes larger. Therefore, it can be concluded that the velocities and trajectories of the particle-laden jets in a crossflow change depending on the solids-gas loading ratio, particle size, carrier gas to free stream velocity ratio and particle to gas velocity ratio at the jet exit.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2016.10a
• /
• pp.609-610
• /
• 2016

본 논문에서는 컨볼루션 신경회로망(CNN: Convolutional Neural Network)과 다양한 분류기들의 결합을 통해 분류성능을 비교하고자 한다. 현재 일반적인 분류기로 알려진 것은 나이브 베이즈(Naive bayes), 트리(Tree), 판별 분석(Discriminant Analysis), 서포트 벡터 머신(SVM: Support Vector Machine) 등이 존재한다. 분류기들은 각각 다른 원리로 분류하기 때문에, 각각 성능을 비교해볼 필요가 있다. 분류기들의 성능을 비교하기 위한 사용한 데이터는 CNN에서 자주 사용되고 있는 MNIST 데이터를 사용하였다. 실험 결과로는 CNN에 선형 SVM을 결합하여 사용한 것이 분류율과 분류속도 측면에서 다른 분류기들의 성능보다 좋은 성능을 보이는 것을 확인할 수 있었다.

• Journal of Internet Computing and Services
• /
• v.12 no.3
• /
• pp.89-99
• /
• 2011

The payload signature-based traffic classification system has to deal with large amount of traffic data, as the number of internet-based applications and network traffic continue to grow. While a number of pattern-matching algorithms have been proposed to improve processing speedin the literature, the performance of pattern matching algorithms is restrictive and depends on the features of its input data. In this paper, we studied how to optimize the search space in order to improve the processing speed of the payload signature-based traffic classification system. Also, the feasibility of our design choices was proved via experimental evaluation on our campus traffic trace.

• 한국HCI학회:학술대회논문집
• /
• 2007.02a
• /
• pp.841-847
• /
• 2007

최근 삶의 수준의 향상과 의학 기술의 발전으로 노인 인구가 증가하고 있다. 하지만 늘어나는 노인 인구에 비례하여 신체적 노화로 거동이 어려운 노인의 수 또한 증가하는 추세이다. 실제로 많은 노인 인구가 거동이 불편해 정상적인 생활을 하지 못하고 있기 때문에 보행 시 적절한 힘을 보조해 줄 수 있는 보행 보조 장치의 개발이 필요하다. 이 같은 보행 보조 장치를 개발함에 있어 보행자의 보행 패턴이 고려된다면 보행자의 걸음걸이에 맞춰 자연스럽게 힘을 보조해 줄 수 있기 때문에 보행자의 보행 단계 분류에 관한 연구가 선행되어야 한다. 그래서 본 논문에서는 하지 근전도 신호를 이용해 보행 단계를 구분하는 방법을 제안하고자 한다. 근전도 신호는 근육이 움직일 때 발생하는 아주 작은 전기적인 신호이다. 근전도 신호는 작은 잡음에도 민감하며, 전극을 부착하는 근육의 위치에 따라서도 값의 차이가 크기 때문에 근전도 신호의 획득 및 처리 방법이 중요하다. 위를 위해 피실험자 별 근육의 위치와 보행 속도를 달리하여 근전도 신호를 획득하고 획득한 신호로부터 여러 특징 값을 추출한다. 그리고 새로운 데이터에 대해 적응성이 강하고 시간에 따라 변하는 근전도 신호의 특성을 잘 반영할 수 있으며 각 집합(class)의 비선형 분리가 가능한 퍼지 최대-최소 신경망(Fuzzy Min-Max Neural Network: FMMNN)을 이용해 보행 단계를 분류해 본다. 실험 결과를 통해 제안한 방법의 타당성을 검증해 보고 보행자, 보행속도, 근전도 측정을 위한 근육의 위치가 보행 패턴 분류에 미치는 영향을 알아본다.

• Archives of design research
• /
• v.14
• /
• pp.267-281
• /
• 1996

One product has so many different kind of user groups and each user group has different characters and needs of use. So, designer has difficulty in definding which group is major or sub in a product, what kind of characters and needs does that group has. In this study, we classify the user upon three factors ; accuracy of rules, speed of action, speed of problem solving and propose design guideline for the classifed user group. The results of this study will be applied to more user-friendly design approach.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.36 no.2C
• /
• pp.78-85
• /
• 2011

In this paper we proposed an intelligent PTZ camera system which detects, classifies and tracks moving objects. If a moving object is detected, features are extracted for classification and then realtime tracking follows. We used GMM for detection followed by shadow removal. Legendre moment is used for classification. Without auto focusing, we can control the PTZ camera movement by using center points of the image and object's direction, distance and velocity. To implement the realtime system, we used TI DM6446 Davinci processor. Throughout the experiment, we obtained system's high performance in classification and tracking both at vehicle's normal and high speed motion.

• Proceedings of the Korean Information Science Society Conference
• /
• 2003.10a
• /
• pp.94-96
• /
• 2003

급속도로 발전하는 인터넷의 발달로 인한 정보의 과부하와 이메일의 급증은 이젠 모든 네티즌들이 겪는 불편함이 아닐 수 없다. 본 논문에서는 이런 이메일 관리를 사용자가 효율적으로 할 수 있도록 추천 에이전트(Recommendation Agent)를 제안하고자 한다. 추천 에이전트 시스템에서는 이메일의 자동 분류에서 가장 핵심인 정확도(Accuracy)를 개선시키기 위해 최종 결정을 사용자가 하는 방식으로 접근하였으며, 또한 절기에 이용되는 학습 및 분류 알고리즘을 동적 임계치를 적용한 베이지안 학습 알고리즘을 이용하여 알고리즘적 방법도 병행하였다. 새로운 메일이 도착했을 때 최적의 분류를 할 수 있도록 메일 카테고리를 추천하는 시스템이다. 또한 사용자 편의를 위하여 필요없는 메일이나 스팸으로 간주되는 메일은 자동 삭제하는 기능을 추가하였다.