基于粗糙集相关矩阵的属性约简算法

An Attribute Reduction Algorithm Based on the Relation Matrix of Rough Set

利用粗糙集相关矩阵采用贪婪策略构造了寻找最小属性约简的启发式算法,证明了算法的正确性并作了复杂性分析,通过实例和与基于属性频率重要性算法进行的对比分析,发现该文算法能快速逼近最小约简,且获得的知识容易理解。

The paper constructed the smallest attribute reduction heuristic algorithm using the relation matrix of rough set and greedy strategy. The correctness and complexity of the algorithm was established. It was discqvered that the algorithm of the paper might approach the smallest attribute reduction apace and prior to the algorithm based on the frequency importance of the attribute.