基于神经网络的遗传算法
【摘要】 随着人工智能技术的发展,遗传算法已经在各个领域得到广泛的应用。然而,随着科技要求的提高,其缺陷也越来越明显。如遗传算法在求解最佳阈值的过程中,需要对每个待进化的个体进行适应度值的计算,这就需要耗费很多的运行时间。为了降低算法的复杂性,笔者提出一种使用神经网络充当算法适应度函数的改进遗传算法。如果设想的方法能够成功,那么我们不仅提供了一种新的适应度函数优化方法,而且能更加真实地反应物种进化的过程。对生物学研究也会有启发。
实验最后表明,本文提出的方法能够达到一定的降低算法复杂性,降低算法对问题的精确性的依赖的目的,为遗传算法提供了一种可行的改进方法,加快了遗传算法运行的速率。
【关键词】 遗传算法,神经网络,适应度函数,BP算法
Genetic algorithm based on neural network
【Abstract】 With the development of artificial intelligence technology, genetic algorithm has been widely used in various fields. However, with the improvement of scientific and technological requirements, its defects are becoming more and more obvious.When the genetic algorithm solves the optimal threshold, need calculate each evolution fitness function. It requires the consumption of a lot of running time. In order to reduce the complexity of algorithm, the author proposes neural network as its fitness function. If the proposal succeeds, we will not only provide a new fitness function optimization method, but also true to reflect about pecies evolution process. Biology research will also be inspired.
Finally,test showed that The proposed method can achieve the purpose of reducing the complexity of the algorithm and reducing the dependence of the algorithm about the accuracy of the problem.It can provide a feasible improvement methods for genetic algorithms, speed up the operation of genetic algorithm.
【Key Words】 genetic algorithm,neural network,fitness function,BP algorithm
图目录
图2.1 简单遗传算法流程图
图2.2 神经网络适应度遗传算法模型
图3.1 神经网络算法的训练
表目录
表4.1 测试函数运行的结果 14