电信欺诈行为的聚类研究

Cluster Analysis of Telecom Fraud

针对海量电信数据的聚类问题,利用粗集中的知识简化方法,减少属性的数量,提取主要的特征属性,并结合性能优良的模糊Kohonen聚类网络,提出了一种新的电信欺诈行为的检测模型,采用Microsoft SQL2005和VC++6.0技术,利用电信运营商提供的真实数据对该模型进行验证,实验结果表明,基于粗集神经网络方法提出的模型快速有效且具有较高的准确率。

This paper combines the rough set theory and neural network for the problem of massive telecom data clustering problem,that is,using the simplified method of rough sets to reduce the number of attributes and extract the main features combining with excellent performance of the fuzzy Kohonen clustering network,a new telecommunications fraud detection model is presented.The model is validated through the data provided by the telecom carriers.The experimental results show that the model is efficient and accurate.