|
Published Articles >> Table of Contents >> Abstract
17th International Conference on Pattern Recognition (ICPR'04) - Volume 1
pp. 148-151
W-Boost and Its Application to Web Image Classification
Jingrui He, Tsinghua University, Beijing, China
Mingjing Li, Microsoft Research Asia, China
Hong-Jiang Zhang, Microsoft Research Asia, China
Changshui Zhang, Tsinghua University, Beijing, China
Full Article Text:
  
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/ICPR.2004.1334029
Send link to a friend
| Abstract |
|
When training data is not sufficient, boosting algorithms tend to overfit as more weak learners are combined to form a strong classifier. In this paper, we propose a new variant of RealBoost, called W-Boost, which is based on a novel weight update scheme and uses changeable bin number to estimate marginal distributions in weak learner design. This new boosting procedure results in both fast convergence rate and small generalization error. Experimental results on synthetic data and web image classification demonstrate the effectiveness of our approach.
|
 Additional Information
|
Citation:
Jingrui He, Mingjing Li, Hong-Jiang Zhang, Changshui Zhang,
"W-Boost and Its Application to Web Image Classification,"
icpr,
pp. 148-151,
17th International Conference on Pattern Recognition (ICPR'04) - Volume 1,
2004
|
|
