摘要： 针对粗糙集属性约简问题，提出一种以最小属性子集和最大依赖度为目标的多目标粒子群优化算法。该算法以非支配排序策略为基础，利用加权法寻找最优粒子，使粒子群在新的运动方程和ε-邻域变化策略的混合作用下进化，从而具有更好的全局开拓和局部收搜索能力。在UCI标准数据集上的对比测试结果表明，该算法具有较好的收敛性。

关键词: 粗糙集, 非支配排序, 属性约简, 多目标进化, 适应度

Abstract: Aiming at the problem of attribute reduction, this paper proposes a new multi-objective particle swarm optimization algorithm which aims at the least reduction of attributes sets and maximum of the dependency of the attributes. It searches best particle based on non-dominated sorting and weighting method. Combining the new motion equation with ε-neighborhood perturbation, the algorithm has strong global and local searching ability. Experimental results with UCI data sets show that the proposed algorithm is more effective than the compared algorithms.

Key words: rough set, non-dominated sorting, attribute reduction, multi-objective evolution, fitness

中图分类号: 

• TP18