Skip to Main Content
This paper proposes a new data table decomposition algorithm to address problems in extracting rules from massive data tables, e.g. low effectiveness, low computing speed and long rule length. With rough set theory, in the perspective of improving classification correctness and sub data table purity, this paper brings up attribute selection measure, and proposes to stop decomposing process to reduce the length of rules. Decompose data table with measures, and finally obtain a rule set with certain supportiveness. The length of obtained rules is short, the effectiveness of decision analysis is high, and it effectively overcomes the impact of noise on rough set data analysis.
Date of Conference: 25-26 Sept. 2008