一种基于粗糙集和遗传神经网络的数据分类器模型

为了实现对无线传感器网络监测得到的高维冗余且不确定的数据进行分类识别,提出一种由遗传算法和粗糙集进行优化的BP神经网络数据分类器模型,并形成了数据挖掘分类算法。该模型通过粗糙集理论的属性约简算法删除训练样本的冗余属性,利用遗传算法对神经网络的权值和阈值进行优化,并进行神经网络学习。数据挖掘分类算法学习速度快,能够有效提高无线传感器网络中数据的分类效率。

A Model of Classifier based on Rough Set and Genetic Neural Network

In order to realize classification and recognition for the high dimensional redundancy and uncertain data monitored by wireless sensor network, a BP neural network data classifier model optimized by genetic algorithm and rough set is proposed, and then a classification algorithm of data mining is formed. In the model, attribute reduction algorithm of rough set theory is used to delete redundant attribute of training samples, and then genetic algorithm is used to optimize weights and threshold values, and carry out neural network learning. The classification algorithm of data mining has better learning speed, and can improve the efficiency of data classification in wireless sensor network.