• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제3권1호
• /
• pp.44-51
• /
• 2003

Knowledge acquisition is a bottleneck in knowledge-based system implementation. Decision tree induction is a useful machine learning approach for extracting classification knowledge from a set of training examples. Many real-world data contain fuzziness due to observation error, uncertainty, subjective judgement, and so on. To cope with this problem of real-world data, there have been some works on fuzzy classification rule learning. This paper makes a survey for the kinds of fuzzy classification rules. In addition, it presents a fuzzy classification rule learning method based on decision tree induction, and shows some experiment results for the method.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2014년도 추계학술발표대회
• /
• pp.936-939
• /
• 2014

전자여권(e-Passport)은 새로운 형태의 출입국 관리 시스템으로써 기존 여권보다 보안 기능이 강화 되고 자동화된 출입국 관리를 할 수 있다는 장점으로 인해 전 세계적으로 도입하기 위한 연구가 활발히 진행되고 있다. 본 연구에서는 전자여권 인식장치로부터 전달 받은 기계판독영역(MRZ : Machine Readable Zone) 이미지의 문자 인식률 향상을 위해 정점 샘플링 방법을 적용하여 문자 인식 결과의 오류를 줄이고자 하였다. 실험 환경에서 오류가 가장 많았던 숫자'1'과 영문'T'에 대해 제안 방법을 적용 하였다. 실험 결과, 제안 방법 적용 전보다 입력 이미지의 모든 문자를 정확히 인식한 이미지의 숫자가 4.5% 증가하였고, 각각의 글자에 대한 전체 오차율은 0.242% 감소하였다.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2012년도 추계학술발표대회
• /
• pp.183-186
• /
• 2012

IT 기술의 급격한 발달로 인해 개인의 생활뿐만 아니라 국내 기업들의 업무 환경까지도 많은 변화가 일어나고 있다. 이러한 변화 중 가상화 기술을 이용하여 업무의 효율성 증대와 경제적이며 관리 능률의 향상을 기대하며 가상화 환경을 도입하려는 기관이 많이 생겨나고 있다. 하지만 가상화 기술은 안정적인 서버의 운영이 뒷받침 되지 않는다면 막대한 피해를 줄 수 있다. 따라서 본 논문에서는 데스크탑 가상화 환경에서 프로비저닝(Provisioning) 과정을 이용하여 가상 서버의 시스템 자원을 최적화시키고, 최적화된 가상 서버에 사용자VM(Virtual Machine)을 할당하는 부하 분산 방안을 제안한다.

• 한국인터넷방송통신학회논문지
• /
• 제16권4호
• /
• pp.147-152
• /
• 2016

데이터에 내포되어 있는 주요 정보나 지식을 추출해 내는 기계학습 방법에서 주요 이슈의 하나는 지식 표현 방식이다. 여러 가지 구조로 표현될 수 있는 지식을 모델이라고 부른다. 모델에는 그 내부 구조에 따라 트리구조, 네트워크 구조, 리스트 구조, 규칙 등 다양한 구조로 나눈다. 구조의 차이는 단지 표현의 차이뿐만 아니라 그것이 갖는 문제해결 능력에도 차이가 있다. 본 논문에서는 모델을 간략화 시켜 오버피팅 문제를 해결하고 분류 능력을 향상시키는 방법을 제안한다. 모델을 단순화 시키는데 사용되는 파티션 유틸리티 기준함수 제시하고 휴리스틱을 이용하여 균형 잡힌 계층 구조를 생성하는 방법을 제안한다.

• 산업경영시스템학회지
• /
• 제27권1호
• /
• pp.1-8
• /
• 2004

This study intend to make easy modification, even if there is a new job or structure change, by modularizing program and computerize and automation of production control management used in CIM. under the condition where manager control production on the job-site, for increasing connection with other operation and management on the computer by monitoring center computer, recognizing information by computer is needed, it is possible by converting transaction. So this study goal is to make delivery control and order control fast and accurate by finding dynamic history of machine and production information in enterprise without input production and quality information by themselves with quality information system. So production increase and quality improvement are possible by diminishing manager's and producer's work with the result of the study combining POP and CIM, after that, in e-business and m-business period that every enterprise must pass, customer satisfaction and sales promotion are possible with employee's computerizing minds. these study result also can knowledge process condition with theoretical class and have a power in finding a solution with foundation of theoretical knowledge.

• 한국경영과학회:학술대회논문집
• /
• 대한산업공학회/한국경영과학회 1994년도 춘계공동학술대회논문집; 창원대학교; 08월 09일 Apr. 1994
• /
• pp.653-659
• /
• 1994

Human experts have the various own knowledges to be applied in specialized domains. The fact that knowledge itself becomes more critical in the context of textile knowledge with rapid development of new fibers, automated equipments, processes and applications. Diversity of worsted spun yarns, lack of human expertise, and inconsistency among manually generated process plans in consequency of adjustment machine parameters owing to change up raw materials frequently increase the necessity of developing computer aided process planning(CAPP) systems for spinning process. Expert systems offer one of techniques to develop CAPP systems which would behave in a knowledgeable manner. Expert systems are the problem-solving computer program that can reach a level of performance comparable to that of a human expert in some specialized problem domain. This paper is described as job justification module. The job justification module performs to consult with users on which worsted spun yarn manufacturing process planning under the various factors, e.g., raw materials, machine parameters and required yarn counts. Also, the developed module informs the various knowledges relevant process planning. The job justification module offers the control parameters at each process and includes the various standard process plans as database. These knowledges are generated by facts and rules within rule bases.

• 한국경영과학회지
• /
• 제22권2호
• /
• pp.79-97
• /
• 1997

Many dispatching rules have been developed for the on-line control of product flow in a job shop. The introduction of a flexible manufacturing system (FMS) has added a new requirement to classical job shop control problem : the selection of machines by parts of different types. An FMS can keep a great deal of information on the status of the system, such as information on what is scheduled in the near future, with great accuracy. For example, the knowledge of the time when the next part will arrive at each machine can be neneficial for the routing. This paper tests the effects of the use of this knowledge for part routing on the parts flow time (sum of the time for waiting and service) under a simple routing procedure- a look-ahead routing procedure. A test under many operating conditions shows that the reduction of part flow time from the cases without using this information is between 1% and 11%, which justifies more study on this routing procedure at real production sites when machine capacity is a critical issue. The test results of this paper are also valid for other highly automated systems such as the semi-conductor fabrication plants for routing when the arrivals of parts in the near future are known.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2000년도 하계학술대회 논문집 D
• /
• pp.2371-2373
• /
• 2000

This paper presents a new vector control scheme for induction motor. An exact knowledge of the rotor flux position is essential for a high-performance vector control. The position of the rotor flux is measured in the direct scheme and estimated in the indirect schemes. Since the estimation of the flux position requires a priori knowledge of the induction motor parameters, the indirect schemes are machine parameter dependent. The rotor and stator resistance among the parameters change with temperature. Variations in the parameters of induction machine cause deterioration of both the steady state and dynamic operation of the induction motor drive. Several methods have presented to minimize the consequences of parameter sensitivity in indirect scheme. In this paper, new estimation scheme of rotor flux position is presented to eliminate sensitivity due to variation in the resistance. The simulation is executed to verify the proposed vector control performance and to compare its performance with that of indirect vector control.

• 한국인터넷방송통신학회논문지
• /
• 제9권4호
• /
• pp.139-144
• /
• 2009

최근의 여러분야에서 검색되어지고 있는 인터넷 도메인 용어의 전문성의 표시화는 온톨리지를 통한 지식의 축적의 목표가 되고 있다. 도메인 용어의 중요성을 표시화 한다면 기계가 온톨리지를 이용하여 정보의 관리 및 해석을 스스로 하는 것이 가능할 것으로 본다. 본 논문에서는 온톨로지의 중요성 (weight)을 구성하는 속성을 확장된 퍼지를 사용하여 기존 웹문서의 구조정보로부터 추출하는 알고리즘을 제안하였다. 특히 속성정보로 구성된 도메인 지식을 표시화 함으로써 속성추출 알고리즘을 개선하고, 추출결과의 품질을 향상시킨다. 5만문서를 대상으로 제안된 알고리즘을 적용한 결과 약 94%의 신뢰도의 속성정보를 추출할 수 있었다.

• ETRI Journal
• /
• 제43권6호
• /
• pp.966-977
• /
• 2021

Filterbank multicarrier with offset quadrature amplitude modulation (FBMC-OQAM) is an attractive alternative to the orthogonal frequency division multiplexing (OFDM) modulation technique. In comparison with OFDM, the FBMC-OQAM signal has better spectral confinement and higher spectral efficiency and tolerance to synchronization errors, primarily due to per-subcarrier filtering using a frequency-time localized prototype filter. However, the filtering process introduces intrinsic interference among the symbols and complicates channel estimation (CE). An efficient way to improve the CE in FBMC-OQAM is using a technique known as windowed frequency domain averaging (FDA); however, it requires a priori knowledge of the window length parameter which is set based on the channel's frequency selectivity (FS). As the channel's FS is not fixed and not a priori known, we propose a k-nearest neighbor-based machine learning algorithm to classify the FS and decide on the FDA's window length. A comparative theoretical analysis of the mean-squared error (MSE) is performed to prove the proposed CE scheme's effectiveness, validated through extensive simulations. The adaptive CE scheme is shown to yield a reduction in CE-MSE and improved bit error rates compared with the popular preamble-based CE schemes for FBMC-OQAM, without a priori knowledge of channel's frequency selectivity.