基于双Kohonen神经网络的Web用户访问模式挖掘算法

本文根据Kohonen自组织特征映射神经网络中学习阶段的性质,运用双Kohonen神经网络组合成新的自组织训练挖掘模型,先使用粗调整训练,加快模型学习速度,紧接着使用微调整训练,提高模型学习精度。实验结果表明,本文提出的双Kohonen神经网络挖掘模型,相对于标准Kohonen神经网络在训练速度和收敛效果上都有一定程度的提高,改善了聚类效果,为挖掘用户的多种兴趣提供了一种可行的方法。

A Web User Access Pattern Mining Algorithm Based on the Dual Kohonen Neural Network

According to the Kohonen self-organizing feature mapping neural network in the nature of the learning stage,the paper uses the dual Kohonen neural network to build a new self-organizing mining model,in order to speed up the learning of the model,utilize the rough adjustment training firstly,and then use the micro-adjustment training for the sake of improving the learning precision of the model.The experimental results show:by contrast to the standard neural network Kohonen,the dual Kohonen neural network,which is written in this article,has been improved in the training speed and the convergence results.It makes the clustering results become much better,and it provides a practicable method to tap the users' various interests as well。