基于改进遗传算法的数据特征分类

Data feature classification based on improved genetic algorithm

针对传统遗传算法在数据特征分类过程中容易陷入局部最佳解,分类结果识别率以及准确率较低的问题,提出基于改进遗传算法的数据特征分类方法。采用模拟退火法对遗传算法实施改进,遗传算法经过设置参数、适应度函数的设计、选择策略、交叉策略以及终止条件等过程得到粗糙数据特征分类结果。采用模拟退火算法通过概率突跳特性在温度下降时随机获取目标函数的全局最优解,基于Meteopolis准则提高算法局部寻优效率,通过模拟退火算法对遗传算法的交叉概率与变异概率的选择过程实施改进,获取高精度的数据特征分类结果。实验结果表明,所提方法数据特征分类识别率以及准确率高,分类耗时低。

As the traditional genetic algorithms may easily fall into the local optimal solution,and has low recognition rate and accuracy rate of classification results during the process of data feature classification,a method of data feature classification based on improved genetic algorithm is proposed. The simulated annealing method is adopted to improve the genetic algorithm which experiences the processes such as parameter setting,fitness function design,selection strategy,crossover strategy,and termination condition,so as to obtain the rough classification result of data features. The simulated annealing algorithm is adopted to randomly obtain the global optimal solution of the objective function by using the probability abrupt-jump feature when the temperature falls,the local optimizing efficiency of the algorithm is improved based on the Meteopolis criterion,and the selection process for crossover probability and mutation probability of the genetic algorithm is improved by means of the simulated annealing algorithm,so as to obtain high-accurate classification result of data features. The experimental results show that the proposed method has high recognition rate and accuracy rate of data feature classification and low classification time consumption.