基于遗传算法和模拟退火算法的特征选择方法

针对模式识别时原始特征数量大而有冗余的现象，提出了一种基于遗传退火算法的特征选优方法。首先对遗传算法和模拟退火做了简要评论，然后在遗传算法中引入模拟退火的Boltzmann更新机制，以克服传统的遗传算法易于过早收敛和易于陷入局部极小的问题。最后阐述、设计了适应度函数和遗传算子。仿真实验表明，该方法在求解的效率和解的质量方面都达到了令人满意的效果。

Feature Selection Method Based on Genetic and Simulated Annealing Algorithm

Aimed to the problem that original feature is mass and redundancy in pattern recognition, a method of featnre optimal based on genetic and simulated annealing algorithm is proposed. This paper firstly describes the genetic algorithm and simulated annealing algorithm, then it introduces the Boltzmann upgrade mechanism into the traditional genetic algorithm to solve the problem of premature convergence of the process and local minima. Finally, it tells and designs a fitness function and genetic operator. Simulations results show that this method has good performance in both the quality of obtained feature subset and efficiency.