Enhanced minimum attribute reduction based on quantum-inspired shuffled frog leaping algorithm  

作　　者：Weiping Ding Jiandong Wang Zhijin Guan Quan Shi 

机构地区：[1]College of Computer Science and Technology,Nanjing University of Aeronautics and Astronautics [2]School of Computer Science and Technology,Nantong University [3]State Key Laboratory for Novel Software Technology,Nanjing University

出　　处：《Journal of Systems Engineering and Electronics》2013年第3期426-434,共9页系统工程与电子技术（英文版）

基　　金：supported by the National Natural Science Foundation of China(61139002;61171132);the Funding of Jiangsu Innovation Program for Graduate Education(CXZZ11 0219);the Natural Science Foundation of Jiangsu Education Department(12KJB520013);the Applying Study Foundation of Nantong(BK2011062);the Open Project Program of State Key Laboratory for Novel Software Technology,Nanjing University(KFKT2012B28);the Natural Science Pre-Research Foundation of Nantong University(12ZY016)

摘　　要：Attribute reduction in the rough set theory is an important feature selection method, but finding a minimum attribute reduction has been proven to be a non-deterministic polynomial （NP）-hard problem. Therefore, it is necessary to investigate some fast and effective approximate algorithms. A novel and enhanced quantum-inspired shuffled frog leaping based minimum attribute reduction algorithm （QSFLAR） is proposed. Evolutionary frogs are represented by multi-state quantum bits, and both quantum rotation gate and quantum mutation operators are used to exploit the mechanisms of frog population diversity and convergence to the global optimum. The decomposed attribute subsets are co-evolved by the elitist frogs with a quantum-inspired shuffled frog leaping algorithm. The experimental results validate the better feasibility and effectiveness of QSFLAR, comparing with some representa- tive algorithms. Therefore, QSFLAR can be considered as a more competitive algorithm on the efficiency and accuracy for minimum attribute reduction.Attribute reduction in the rough set theory is an important feature selection method, but finding a minimum attribute reduction has been proven to be a non-deterministic polynomial （NP）-hard problem. Therefore, it is necessary to investigate some fast and effective approximate algorithms. A novel and enhanced quantum-inspired shuffled frog leaping based minimum attribute reduction algorithm （QSFLAR） is proposed. Evolutionary frogs are represented by multi-state quantum bits, and both quantum rotation gate and quantum mutation operators are used to exploit the mechanisms of frog population diversity and convergence to the global optimum. The decomposed attribute subsets are co-evolved by the elitist frogs with a quantum-inspired shuffled frog leaping algorithm. The experimental results validate the better feasibility and effectiveness of QSFLAR, comparing with some representa- tive algorithms. Therefore, QSFLAR can be considered as a more competitive algorithm on the efficiency and accuracy for minimum attribute reduction.

关 键 词：minimum attribute reduction quantum-inspired shuf- fled frog leaping algorithm multi-state quantum bit quantum rotation gate and quantum mutation elitist frog. 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TP38[自动化与计算机技术—控制科学与工程]