NIPS
2004
14 years 28 days ago
2004
In this paper we propose to combine two powerful ideas, boosting and manifold learning. On the one hand, we improve ADABOOST by incorporating knowledge on the structure of the dat...