遗传神经网络在邮件过滤器中的应用

ANNA Optimized by GA and Its Application in E-mail Filter

针对目前反垃圾邮件技术的缺点,提出一种基于遗传优化神经网络的垃圾邮件过滤器模型,利用遗传算法全局搜索能力优化神经网络连接权值,克服神经网络局部极小值点问题,提高神经网络的学习速度和识别能力。通过对遗传算法和人工神经网络算法的实现,证明它们在垃圾邮件过滤器中有很好的应用效果。

This paper presents a model of spare mail filter based on artificial neural network which is optimized by genetic algorithms. By using genetic algorithms, which is good at wide searching ability in solution space on optimizing connection weight matrix of artificial neural network, artificial neural network can get over the inherent limitation of local minimal and improve its learning speed and recognition ability. The application of the model in spare filter has been successfully proved by its implementing.