Home

About

Advanced Search

Browse by Discipline

Scientific Societies

E-print Alerts

Add E-prints

E-print Network
FAQHELPSITE MAPCONTACT US


  Advanced Search  

 
Multiclass Boosting with Repartitioning Ling Li ling@caltech.edu
 

Summary: Multiclass Boosting with Repartitioning
Ling Li ling@caltech.edu
Learning Systems Group, California Institute of Technology, Pasadena, CA 91125, USA
Abstract
A multiclass classification problem can be re­
duced to a collection of binary problems with
the aid of a coding matrix. The quality of the
final solution, which is an ensemble of base
classifiers learned on the binary problems, is
a#ected by both the performance of the base
learner and the error­correcting ability of the
coding matrix. A coding matrix with strong
error­correcting ability may not be overall op­
timal if the binary problems are too hard for
the base learner. Thus a trade­o# between
error­correcting and base learning should be
sought. In this paper, we propose a new mul­
ticlass boosting algorithm that modifies the
coding matrix according to the learning abil­
ity of the base learner. We show experimen­

  

Source: Abu-Mostafa, Yaser S. - Department of Mechanical Engineering & Computer Science Department, California Institute of Technology

 

Collections: Computer Technologies and Information Sciences