Seminar: Classification using One-class SVM under Lable Noise

Thesis Proposal Presentation

Hao Yuan
M.Sc. Thesis Proposal
Co-Supervisors: Dr. Jian Tang & Dr. Minglun Gong

Classification using One-class SVM under Lable Noise

Department of Computer Science
Friday, November 22, 2013, 11:40a.m., Room EN 2022


The presence of noise in training set has strong impact on the performance of supervised learning (classification) techniques. In classification area, there are two types of noise: feature noise and label noise. Modeling and dealing with these noises are important; however, there is not much literature
related to label noise. As the one-class SVM algorithm has advantage over standard binary SVM on handling ambiguous data, there is a strong indication that using One-class SVM can better handle label noise. In the research, we will analysis and model label noise, design one-class SVM based approach to deal with label noise, and try to apply this method to social network classification.