【摘要】 基于RoughSet理论,研究从Skrowon分辨矩阵到规则自动生成过程中的关键算法的改进问题。提出由分辨矩阵到合取项矩阵的计算方法,建立了从属性约简的合取项矩阵到析取项矩阵转换的数学模型,基于数学模型,提出直接搜索的转换方法。同时,提出了属性值约简的改进方法,使得改进后的算法从总体上节省了运算空间,降低了算法的时间复杂性,提高了规则生成的效率。最后通过UCI数据库的实例验证了改进算法的有效性。更多还原

【Abstract】 The automatic rule generation algorithms from Skrowon discernibility matrix are studied and improved respectively based on Rough Set Theory.The tansformation method form discernibility matrix to conjunction matrix is presented.And the mathimetic model of the transformation from conjunction matrix to disjunction matrix is esteblished for the attribute reduct.Based on the model,direct search method is presented.Also,the attribute value reduction algorithm is impoved.By this means,the whole rule ganeration can save the computing space and CPU occupation time,enhance the efficiency of decision rules generation.Finally,by the factual process for UCI database,the validity of two improved algorithms are proved.更多还原