brianwu-s / domain_adaptation Goto Github PK
View Code? Open in Web Editor NEWDomain Adaptation is one of the branches of migration learning. Its purpose is to map differently distributed source and target domain data into a feature space, minimizing the distance between the two in that space. In this report, we use Office-Home dataset as our dataset, and try traditional methods (TCA, JDA, CORAL, BDA, WBDA) and deep learning methods (DANN, DDC DeepCORAL) to do domain adaptation and evaluate the performances of them.