• 제목/요약/키워드: Datamining Roadmap

검색결과 2건 처리시간 0.013초

Datamining: Roadmap to Extract Inference Rules and Design Data Models from Process Data of Industrial Applications

  • Bae Hyeon;Kim Youn-Tae;Kim Sung-Shin;Vachtsevanos George J.
    • International Journal of Fuzzy Logic and Intelligent Systems
    • /
    • 제5권3호
    • /
    • pp.200-205
    • /
    • 2005
  • The objectives of this study were to introduce the easiest and most proper applications of datamining in industrial processes. Applying datamining in manufacturing is very different from applying it in marketing. Misapplication of datamining in manufacturing system results in significant problems. Therefore, it is very important to determine the best procedure and technique in advance. In previous studies, related literature has been introduced, but there has not been much description of datamining applications. Research has not often referred to descriptions of particular examples dealing with application problems in manufacturing. In this study, a datamining roadmap was proposed to support datamining applications for industrial processes. The roadmap was classified into three stages, and each stage was categorized into reasonable classes according to the datamining purposed. Each category includes representative techniques for datamining that have been broadly applied over decades. Those techniques differ according to developers and application purposes; however, in this paper, exemplary methods are described. Based on the datamining roadmap, nonexperts can determine procedures and techniques for datamining in their applications.

데이터마이닝 로드맵 개발과 수처리 응집제 제어를 위한 데이터마이닝 적용 (Development of Datamining Roadmap and Its Application to Water Treatment Plant for Coagulant Control)

  • 배현;김성신;김예진
    • 한국정보통신학회논문지
    • /
    • 제9권7호
    • /
    • pp.1582-1587
    • /
    • 2005
  • 본 논문은 정수장에서 사용하는 응집제의 종류를 결정하기 위한 시스템 개발에 관한 내용이다. 정수장은 여러 단위 처리장으로 구성되며, 불순물을 제거하기 위하여 혼화지에서 응집제를 주입하여 침전을 시킨다. 현재까지 응집제 결정을 위해 Jar-test를 이용하는데, 이 방법은 사람의 주관적인 판단에 의존하므로 실험 오차가 발생할 수 있다. 특히 정수장의 자동화를 위한 시스템 개발에서 가장 큰 걸림돌로 작용하고 있다. 본 논문은 이러한 문제점을 해결하기 위하여 로드맵에 기초한 데이터마이닝 기법을 이용하여 응집제를 선택할 수 있는 제어기를 개발하였다. 제어 규칙은 클러스터링 기법으로 도출하였는데, 군집의 초기 값과 개수는 통계적 지수 값을 사용하여 결정하였다.