An Improved Attribute Reduction Algorithm based on Granular Computing

Granular computing is a new intelligent computing method based on problem solving, information processing and pattern classification. Granular com- puting based attribute reduction method is an important application of Granular computing. These algorithms are mostly based on reduction core. However, some information systems may have no reduction core, especially in the actual application data. For this case, those algorithms are powerless. In this paper, an improved reduc- tion algorithm based on granular computing is proposed. The algorithm is validated by the experimental result.