Theory and Applications of Boosting
Published on Jul 30, 200924201 Views
Boosting is a general method for producing a very accurate classification rule by combining rough and moderately inaccurate "rules of thumb". While rooted in a theoretical framework of machine learnin