Abstract: Learning from label proportions (LLP) is a promising weakly supervised learning problem. In LLP, a set of instances (bag) has label proportions, but no instance-level labels are given. LLP ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results