Toward Online Learning on Streaming Data
报告题目:Toward Online Learning on Streaming Data
报告时间:1月19号下午4点
报告地点:工商楼200-9室
Abstract: The proliferation of streaming data in various disciplines motivates a surge of research interests on stream data mining and online learning. First, based on kernel trick we present some online learning algorithms for regression/classification task on streaming data. The meet of class imbalance and concept drift makes the learning problem even harder. We propose an algorithm to exploit the dynamic conditional distribution and imbalance ratio information lurked in data stream. At last, we study how to perform online node classification on streaming networks.
报告人:渐令,男,中国石油大学(华东)理学院副教授、统计学硕士生导师,IEEE会员,SIAM会员教育部学位中心通讯评议专家,国家自然科学基金通讯评审专家。目前主要从事机器学习与数据挖掘及其应用等方面的研究。主持国家自然科学基金(青年、天元青年)、山东省自然科学基金等科研项目。在IEEE Transactions on Neural Networks and Learning Systems、Data Ming and Knowledge Discovery 等期刊和国际会议IJCAI、SDM、BIBM 发表学术论文30 余篇。2014年入选中国石油大学(华东)“青年骨干教师建设工程”人才计划。
联系人:郜传厚(gaochou@zju.edu.cn)
欢迎大家参加!