Learning to Classify with Missing and Corrupted Features
Published on Aug 07, 20083713 Views
After a classifier is trained using a machine learning algorithm and put to use in a real world system, it often faces noise which did not appear in the training data. Particularly, some subset of fea