• DocumentCode
    3371735
  • Title

    Algorithm for decision rules reduction in incomplete information system based on binary discernibility matrix

  • Author

    Bai, Xiuling ; Zhang, MingChuan ; Qiu, Yong ; Wu, Qingtao

  • Author_Institution
    Coll. of Electron. & Inf. Eng., Henan Univ. of Sci. & Technol., Luoyang, China
  • fYear
    2009
  • fDate
    9-12 Aug. 2009
  • Firstpage
    4061
  • Lastpage
    4066
  • Abstract
    Being an important topic of the knowledge reduction of rough set, the reduction of the decision rules respectively eliminates the redundant attribute values of each decision rule based on rough logic. The definition of the binary discernibility matrix in the incomplete decision system was given by analyzing all kinds of definitions of the binary discernibility matrix. An algorithm for decision rules reduction in incomplete information system was designed based on binary discernibility matrix. This algorithm is suitable for both compatible and incompatible decision table. The correctness of the proposed algorithm was proved by examples and the efficiency was indicated by empirical analysis and applications.
  • Keywords
    decision tables; formal logic; matrix algebra; rough set theory; binary discernibility matrix; decision rules reduction; incompatible decision table; incomplete decision system; incomplete information system; knowledge reduction; redundant attribute values; rough logic; rough set; Algorithm design and analysis; Automation; Data mining; Educational institutions; Information analysis; Information systems; Knowledge engineering; Logic; Mechatronics; Set theory; Binary Discernibility Matrix; Decision Rules Reduction; Generalized Decision System; Rough Set;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Mechatronics and Automation, 2009. ICMA 2009. International Conference on
  • Conference_Location
    Changchun
  • Print_ISBN
    978-1-4244-2692-8
  • Electronic_ISBN
    978-1-4244-2693-5
  • Type

    conf

  • DOI
    10.1109/ICMA.2009.5246634
  • Filename
    5246634