Select representative data to train SVM for big data classification


主题:   Select representative data to train SVM for big data classification主讲人:   Xiaoou Li 地点:   松江校区一号学院楼140报告厅时间:   2018-11-15 13:00:00组织单位:   计算机学院


Prof. Xiaoou Li obtained B. S degree of applied mathematics in 1991 and PhD degree of Automatic Control in 1995 from Northeastern University, Shenyang, P. R. China. She has been a professor of Department ofComputer Science, The Research and Advanced Studies Centre of the National Polytechnic Institute (CINVESTAV-IPN), Mexico. She was a senior research fellowof School of Electronics, Electrical Engineering & Computer Science,Queen's University Belfast, UK during the school year 2006-2007 (sabbaticalleave); and school of Engineering, University of California Santa Cruz in 2010(sabbatical leave). Currently she is a senior member of IEEE, member of AMC(Mexican Association of Science), and member of SNI (National ResearcherSystem) level 2. Dr. Li has published more than 100 papers on international journals, book chapters and conferences. She has successfully finished three CONACYT (NSF in Mexico) projects in the field of Knowledge and Data Engineering, and one collaborative project with University of California Riverside.


Support Vector Machines (SVM) has demonstrated highly competitive performance in many real-world applications. However, despite its good theoretical foundations and generalization performance, SVM is not suitable for classifying large data sets because of high training complexity. In recent years, we have introduced several data reduction techniques into SVM classification process to handle this problem, such as minimum enclosure ball, clustering, convex-concave hull, decision tree, etc. Experiments showed that SVM is still suitable for large data classification if the training data is sufficient representative. In this talk, I will present several data selection or reduction techniques and correspondingtraining process.   


视频:   摄影: 撰写:李倩  信息员:李倩  编辑:向娟