Domain Adaptation with Ensemble of Feature Groups

  • Rajhans Samdani ,
  • Scott Wen-tau Yih

Proceedings of the Twenty-Second International Joint Conference on Artificial Intelligence |

Publication

We present a novel approach for domain adaptation based on feature grouping and re-weighting. Our algorithm operates by creating an ensemble of multiple classifiers, where each classifier is trained on one particular feature group. Faced with the distribution change involved in domain change, different feature groups exhibit different cross-domain prediction abilities. Herein, ensemble models provide us the flexibility of tuning the weights of corresponding classifiers in order to adapt to the new domain. Our approach is supported by a solid theoretical analysis based on the expressiveness of ensemble classifiers, which allows trading-off errors across source and target domains. Moreover, experimental results on sentiment classification and spam detection show that our approach not only outperforms the baseline method, but is also superior to other state-of-the-art methods.