Sciweavers

KDD
2002
ACM
157views Data Mining» more  KDD 2002»
14 years 12 months ago
Exploiting unlabeled data in ensemble methods
An adaptive semi-supervised ensemble method, ASSEMBLE, is proposed that constructs classification ensembles based on both labeled and unlabeled data. ASSEMBLE alternates between a...
Kristin P. Bennett, Ayhan Demiriz, Richard Maclin