博彩平台
学术报告
报告题目： A Duality View of Boosting Algorithms and Applications
演讲人： Dr. Chunhua Shen,
报告地点：校本部理工楼504会议室
报告时间：
报告摘要:Consideration of the primal and dual problems together leads to important new insights into the characteristics of boosting algorithms. We show that the Lagrange dual problems of AdaBoost, LogitBoost and soft-margin LPBoost with generalized hinge loss are all entropy maximization problems. By looking at the dual problems of these boosting algorithms, we show that the success of boosting algorithms can be understood in terms of maintaining a better margin distribution by maximizing margins and at the same time controlling the margin variance. We also theoretically prove that, approximately, AdaBoost maximizes the average margin, instead of the minimum margin.The duality formulation also enables us to develop column generation based optimization algorithms, which are totally corrective.we then propose a general framework that can be used to design new boosting algorithms.We show that the proposed boosting framework, termed CGBoost, can accommodate various loss functions and different regularizers in a totally-corrective optimization fashion. We also demonstrate that many boosting algorithms like AdaBoost can be interpreted in our framework--even if their optimization is not totally corrective.Based on this framework, new boosting algorithms are designed and tailored to real-time visual object detection.State-of-the-art performance is achieved in problems like face detection and human detection.