文献类型：期刊文献

英文题名：Prosvms based diagnostic model of chronic gastritis in TCM

作者：Yan, Jian-Jun[1]; Zhong, Tao[1]; Liu, Guo-Ping[2]; Wang, Yi-Qin[2]; Guo, Rui[2]; Zheng, Wu[2]; Qian, Peng[2]

机构：[1] Center for Mechatronics Engineering, East China University of Science and Technology, Shanghai, 200237, China; [2] Laboratory of Information Access and Synthesis of TCM Four Diagnosis, Basic Medical College, Shanghai University of Traditional Chinese Medicine, Shanghai 201203, China

年份：2014

起止页码：1114

外文期刊名：IEEE International Conference on Control and Automation, ICCA

收录：EI(收录号：20143518117248)

语种：英文

外文关键词：Diagnosis - Learning systems

摘要：Multi-label learning task is using to solve problems of syndrome diagnosis for patients may simultaneously have more than one syndrome in traditional Chinese medicine (TCM). The two goals of multi-label learning are label prediction loss and relevance ordering loss. Most Multi-label learning algorithms focus on only one of the goals and neglect the other one. However, there is a multi-label learning algorithm named ProSVMs give consideration to both. And it is apply to the diagnosis of chronic gastritis (CG) of TCM. While its performance suffers from irrelevances and redundancies of the overall feature space of low predict accuracy. Feature selection is combined with ProSVMs to establish the classification model for CG. The result shows the satisfied performance of the diagnostic model for CG was achieved. ? 2014 IEEE.