Evaluation of Classifers using Resampling Methods
Advisor: Christoph Lehmann
Evaluation of a classifier by means of resampling methods like bootstrapping, cross-validation, jack knife, etc. is well established in statistics. Since these methods are computational intensive, they are not yet broadly used within the machine learning community. Because of its structure, the resampling task could benefit from parallelization. The target of this thesis is to elaborate approaches for parallelizing the resampling task in order make it fast and efficient.