现代电子技术2016，Vol.39Issue(9)：37-40,4.DOI:10.16652/j.issn.1004-373x.2016.09.008
利用深度置信网络的中文短信分类
Chinese SMS classification with deep belief nets
摘要
Abstract
To improve the filtering effect of spam SMS,a feature extraction algorithm is proposed to convert SMS content in⁃to fixed length vector with word2vec tool by the analysis of Chinese SMS content and structure characteristics. The deep belief nets(DBN)were designed to learn and classify. The experimental results show that the generalization performance is increased by about 5% in comparison with the reported results.关键词
深度置信网络/深度学习/受限波尔兹曼机/短信Key words
deep belief net/deep learning/restricted Boltzmann machine/SMS分类
信息技术与安全科学引用本文复制引用
王贵新,郑孝宗,张浩然,张小川..利用深度置信网络的中文短信分类[J].现代电子技术,2016,39(9):37-40,4.基金项目
国家自然科学基金项目（60443004）人工生命动觉智能行为选择图式理论研究 （）
