PAC Learning under Helpful DistributionsFrançois Denis and Rémi GilleronRAIRO-Theor. Inf. Appl., 35 2 (2001) 129-148DOI: https://doi.org/10.1051/ita:2001112
