A method for estimating feature weights in CBR
Ruchuan Wang · Journal of Xi'an University of Post and Telecommunications · 2007
One of the key problems of similar matching in CBR is to determine the weights of case features.In this paper,a new method is proposed to estimate the weights of case features.This method applies the Rough Set theory to decide the significance of each feature,and adopts the speedy discretization algorithm to process the continuous features.The experiment results have indicated that the method can make the similar matching results objectively reflecting the similarity between the cases and improve the accuracy of case retrieve.