论文部分内容阅读
研究基于二进制粒子群优化算法思想求解决策表最小属性约简问题的方法.定义适当的适应值函数,将决策表最小属性约简问题转化为一个适合二进制粒子群优化算法求解的0—1组合优化问题,证明问题解的等价性.在此基础上,引入种子粒子概念及其自适应保护策略,提出一个改进的二进制粒子群算法,取得良好的效果.实验结果说明该算法的有效性.
A method based on binary particle swarm optimization algorithm to solve the problem of minimum attribute reduction in decision table is studied.A suitable fitness function is defined and the minimum attribute reduction problem in decision table is transformed into a 0-1 combinatorial optimization suitable for binary particle swarm optimization Problem, and prove the equivalence of the solutions.On this basis, we introduce the concept of seed particle and its adaptive protection strategy, and propose an improved binary particle swarm optimization algorithm with good results. Experimental results show the effectiveness of the algorithm.