报告题目:Unsupervised Ensemble Learning and Its Application to Temporal Data Mining
报告时间:2017年10月20日10:00
报告地点:B403
报告人:杨云教授
报告人单位:云南大学
报告人简介:
杨云,云南大学软件学院教授,博士生导师,云南省中青年学术和技术带头人后备人才,云南省 “百人计划”获得者。2011 年获得英国曼彻斯特大学计算机科学博士学位。攻读博士学位期间,其入选英国政府资助的海外研究生奖励计划(UK Overseas Research Students Awards Scheme –ORSAS)。博士毕业之后,其在英国萨里大学从事研究员工作,期间主持了由欧共体第七个框架计划资助下的国际合作项目的子项目。2014年杨云教授以引进人才形式进入云南大学工作,之后分别主持国家自然科学基金项目2项,参与国家自然科学基金项目1项,主持省部级科研项目4项。2016年杨云教授获多家欧洲高水平科研团队联合邀请作为中方团队负责人,参与欧盟地平线2020国际合作项目(大健康主题)的申请工作。杨云教授的研究方向包括:机器学习,数据挖掘,模式识别,大数据处理与分析等,其代表性研究成果分别发表于《IEEE Transactions on Neural Networks and Learning Systems》,《IEEE Transactions on Knowledge and Data Engineering》,《IEEE Transactions on Systems, Man, and Cybernetics》等国际知名学术期刊,2016年其在国际权威学术出版社 Elsevier 出版全英文学术专著1部。此外,杨云教授进一步把理论研究与实际应用相结合,与多家企事业单位进行了全面的产学研合作,获得多项发明专利与软件著作权,取得了良好的经济与社会效益。
摘要:Ensemble learning is originally proposed for classification tasks in a manner ofsupervised learning,the basic concept of ensemble learningis to train multiple base learners as ensemble members and combine their predictionsinto a single output that should have better performance on average than any other ensemble member. Recently, ensemble learning has been extended to unsupervised learning with different strategies, named as clustering/unsupervised ensemble. This has led to many real-world applications,such as gene classification, image segmentation, video retrieval, and so on. This presentation is going to concentrate on unsupervised ensemble learning technique, and talk about how such technique can deal with the challenge in clustering temporal data with various and high dimensionality, largevolume, very high-feature correlation, and a substantial amount of noise.
邀请人:李元香教授,何国良副教授