一种基于神经网络的垃圾邮件过滤方法

垃圾邮件问题日益严重,受到研究人员的广泛关注,基于各种技术的垃圾邮件过滤方法应运而生,其中神经网络技术应用广泛.现在主要采用的后向传播(BP)神经网络虽然在垃圾邮件过滤中取得很好的效果,但仍然存在局部极小点、不能适应新样本、学习效率较低等诸多问题.因此,本文将一种有导师、可在线学习的自组织神经网络--预测自适应谐振理论神经网络(ARTMAP),运用于垃圾邮件过滤,提出了一种新的基于ARTMAP的垃圾邮件过滤方法.实验表明,基于ARTMAP的邮件过滤能够对垃圾邮件进行有效的过滤,在保证正确率的同时,更能适应当前垃圾邮件特征不断变化的环境.

A Neural Network Based Spam Filtering Approach

The volume of spam in Internet has grown tremendously in the past few years.And this problem has aroused concern from many researchers and they have developed various methods to filter sparn.With the development of neural network and artificial intelligence,neural network has been increasingly used in filtering spam,especially BP neural network.It has achieved good results,but there still exist many problems such as easily stacking into local minimum value,inability to meet sample and low learning efficienc...