基于粗糙集和神经网络的数据挖掘应用

Data Mining Applications Based on Rough Set and Neural Network

本论文就是根据电信行业需求,针对电信企业拥有大量详实而且丰富的数据,但是可用有效数据提取困难这一问题。首先利用粗糙集理论中的差别矩阵方法对电信客户数据进行属性约简,之后采用BP(Back Propagation)神经网络建立基于粗糙集和神经网络的数据挖掘模型,实现对电信业务系统的客户数据信息进行有效分析和高效提取,并通过matlab实现了仿真模拟。所建立的模型,减少神经网络的输入层个数、简化了运算次数、缩短了训练时间并提高数据预测的准确度。

Based on the needs of the telecommunications industry, for telecommunications companies which have a large number of detailed and rich data, but it is difficuh for the extraction of available valid data. First, this paper carriedout attribute reduction to telecommunications customer data using the difference matrix method of rough set theory, and then established data miningmodel based on rough set and neural network, using BP （Back Propagation） neural network, to achieve effective analysis and efficient extraction to customer data information of telecommunication services system, and realize simulationby matlab. The established model reduces the number of the input layer of theneural network, and simplifies the number of operations, shortens the training time and improves the accuracy of the data forecast.