RankSim: Ranking Similarity Regularization for Deep Imbalanced Regression. (arXiv:2205.15236v2 [cs.LG] UPDATED)

Data imbalance, in which a plurality of the data samples come from a small
proportion of labels, poses a challenge in training deep neural networks.
Unlike classification, in regression the labels are continuous, potentially
boundless, and form a natural ordering. These distinct features of regression
call for new techniques that leverage the additional information encoded in
label-space relationships. This paper presents the RankSim (ranking similarity)
regularizer for deep imbalanced regression, which encodes an inductive bias
that samples that are closer …

arxiv lg ranking regression regularization