In this paper, a clustering algorithm on high-dimension data partitional mended attribute is put foreword. First of all, through partition of the attribute value and discriminant degree of element, Conjunctive discriminant rules on training set are got. secondly, different discriminant rules on base of mended attribute are found. Experimental results show: the judgement rules got through the training set can better discriminate the test set. thereby, Experimental results verify the effectiveness of the proposed algorithm.
展开▼