摘要

To the shortcomings of the improved harmony search (IHS) algorithm, The adaptive harmony search (AHS) algorithm is proposed, which uses the maximum difference of objective function values in the harmony memory to adjust PAR and bw, so as to improve the search efficiency of multidimension problems. Then, the proposed AHS algorithm is tested with five criteria, and is applied to relative attribute reduction. The simulation result shows the effectiveness and feasibility of the algorithm.

全文