基于神经网络物联网实体信息交互关系识别

交互关系识别的方法。文中详细的描述了实体交互关系识别模型的设计,并且通过实验表明,逻辑回归受限于其线性分类器的本质在实验结果中稍差,而神经网络通过隐含层中特征空间的转换,在本应用中要优于所有其它的算法Р 参考文献: Р [1] Xia Feng. of things[J]. International Journal munication Systems 25, 2012(9): 1101. Р [2] 孙其博. 物联网: 概念, 架构与关键技术研究综述[J]. 北京邮电大学学报, 2010, 33(3): 1-9. Р [3] Choudhury, Tanzeem Khalid. Sensing and Modeling works[J]. Massachusetts Institute of Technology, 2004: 101-105. Р [4] Eagle N, Pentland A. Reality Mining: Sensing Co mplex Social Systems[J]. Personal and puting, 2006, 10: 255-268. Р [5] Eagle N, Pentland A, Lazer D. Inferring work Structure using Mobile Phone Data[J]. Proceedings of the National Academy of Sciences (PNAS), 2009, 106(36): 15274-15278. Р [6] Onnela J. Structure and Tie Strengths in works[J]. Proceedings of the National Academy of Sciences (PNAS), 2007, 104: 7332