决策强度的决策表约简设计与比较

Design and Comparison of Decision Table Reduction Based on Decision Power

在平均决策强度的基础上,提出了决策强度的代数定义,以弥补基于正区域方法的局限性,并且证明了知识的决策强度随着信息粒度变小而非单调递减的规律,在此基础上设计了基于决策强度的约简方法。应用实例分析的结果表明,基于决策强度的属性重要性是一种更优的启发式信息,该方法计算直观有效,时间复杂度较低,有助于搜索最优或次优约简。最后对UCI离散数据集进行约简比较。

Attribute reduction is one of the important topics in the research on rough set theory. To eliminate the limitations of classical rough reduction algorithms based on positive region, a new decision power definition of algebra was proposed, which was based on the recent mean decision power, and the new significance of an attribute was defined. The conclusion that decision power of knowledge de- creases non - monotonously as the information granularities become finer was obtained, and a heuristic algorithm was proposed. Theo- retical analyses show that the proposed heuristic information is better and more efficient than. the others, and this computation is direct and efficient, and its time complexity is relatively less. Experimental results prove the validity of the heuristic algorithm in searching the minimal or optimal reduction. Finally, the reduction comparison results of UCI discrete databases using four algorithms were gotten.