摘要
IPTV视频业务的复杂性和多样性使其难以充分发挥运营商技术优势。借助深度神经网络DNN模型对IPTV视频用户进行用户行为分析。利用深度学习算法对用户点播视频活跃度实施精确分类,从而帮助IPTV服务提供商合理配置资源,同时为终端用户提供更高效优质的服务。实验结果表明,与现有的方法相比,该方法收敛快,分类准确率达93%。
In view of the complexity and diversity of IPTV video services,it is difficult to give full play to the technical advantages of operators.In this paper,the user behavior of IPTV video users was analyzed by means of deep neural network model.The deep learning method was used to accurately classify the user s on-demand video activity,thereby helping the IPTV service provider to properly allocate resources and provide more efficient and high-quality services for the end users.The experimental results show that compared with the existing methods,the method converges quickly and the classification accuracy rate is 93%.
作者
刘超
贾毓臻
王攀
Liu Chao;Jia Yuzhen;Wang Pan(School of Electrical and Information Engineering, Jiangsu University, Zhenjiang 212013, Jiangsu, China;Modern Postal Research Institute, Nanjing Posts and Telecommunications University, Nanjing 210003, Jiangsu, China)
出处
《计算机应用与软件》
北大核心
2019年第6期167-170,286,共5页
Computer Applications and Software
基金
江苏省博士后基金项目(1402095C)
江苏大学高级人才科研启动基金项目(1291140025)
