Classification is a fundamental task in statistical machine learning, with numerous practical applications in many fields. Classification using boosting (and trees) is particularly popular in industry. This talk will present the general idea of ABC-Boost (adaptive base class boosting) for multi-class classification and its two specific implementations named ABC-MART and ABC-LogitBoost. The original MART (Friedman 2001) and LogitBoost (Friedman, Hastie, and Tibshirani, 2000) algorithms are highly influential in the field of statistical machine learning. We show that, on many public datasets, ABC-MART could improve MART roughly by 10% (relatively) in terms of the mis-classification errors. Furthermore, ABC-Logitboost could improve ABC-MART roughly by another 10%. Bio: Ping Li is an assistant professor in the Department of Statistical Science at Cornell University. In 2007, he graduated from Stanford University, with a Ph.D. in Statistics, an M.S. in Computer Science, and an M.S. in Electrical Engineering. His research interests include (1) fundamental randomized algorithms for processing massive (and possibly streaming) datasets; (2) statistical machine learning. Ping Li is a recipient of the ONR (Office of Naval Research) Young Investigator Award in 2009. UCLA Department of Statistics 2009-09-17T11:03:37-07:00 2009-10-20 Statistical Science true 4:00 PM 419 4660 Geology Bldg. Cornell University false true UCLA Department of Statistics Seminar Ping Li 3:00 PM 2009-10-14T07:25:19-07:00