发帖
雷达卡
Heterogeneous Risk Minimization
Jiashuo Liu 1 Zheyuan Hu 1 Peng Cui 1 Bo Li 2 Zheyan Shen 1
Abstract with ERM usually suffer from poor generalization perfor-
mance due to the greedy exploitation of correlations among
Machine learning algorithms with empirical risk
the training data, which are not stable under distributional
minimization usually suffer from poor generaliza-
...
京公网安备 11010802022788号
