Learning, compression, and leakage: Minimising classification error via meta-universal compression principles.

Fernando E. Rosas, Pedro A. M. Mediano, Michael Gastpar

Browse the full ITW paper archive.