CAJ | 학술논문

为解决传统遗传算法在属性约简时会出现迭代次数多，收敛较慢的问题，论文提出了一种改进的遗传算法。该方法在适应函数上加入属性重要度因子，同时在交叉操作中有选择地保留子代个体，确保算法能够快速收敛。实验结果证明，改进之后的算法在保证属性约简的基础上，能够实现比传统遗传算法更快的迭代和收敛。
위해결전통유전산법재속성약간시회출현질대차수다，수렴교만적문제，논문제출료일충개진적유전산법。해방법재괄응함수상가입속성중요도인자，동시재교차조작중유선택지보류자대개체，학보산법능구쾌속수렴。실험결과증명，개진지후적산법재보증속성약간적기출상，능구실현비전통유전산법경쾌적질대화수렴。
In order to solve the question of more iterative times and the slower convergence speed of traditional genetic algorithm in attribute reduction ,an improved genetic algorithm is proposed .This algorithm gives sufficiency function added a factor of attribute importance ,and saves the individual which have good adaptation in crossover operation ,to make sure the algorithm can quickly converge .The result shows that this method not only ensures attribution reduction ,but also achieves fast iterative and fast convergence .