A Method of Weight Computing in the Comprehensive Estimation

LIU Yu,QIU Yan-zhang,LI Chang-biao
DOI: https://doi.org/10.3969/j.issn.1000-7180.2006.06.024
2006-01-01
Abstract:To determine the attribute significance is an important process in the comprehensive estimation. For some intrinsic shortcomings in traditional methods, the attribute significance is defined scientificly on the base of two approximate quality of rough set, which is as a objective weights in comprehensive estimation according to the thought of information granularity. Because of the experience knowledge factor, the comprehensive weights in favor of right estimation and decision are obtained by using subjective weights and objective weights. The actual application in flow pattern recognition shows the method is complete, and the application effect is superior to that of other methods.
What problem does this paper attempt to address?