ICML
2007
IEEE
15 years 13 days ago
2007
IEEE
In supervised learning, we commonly assume that training and test data are sampled from the same distribution. However, this assumption can be violated in practice and then standa...