On the almost optimality of blind active sampling.
Résumé
We show that blind active sampling is almost optimal in the case of large covering numbers in regression with finite precision. THIS IS A NON-COPYRIGHTED PUBLICATION IN A NATIONAL CONFERENCE.