nips nips2010 nips2010-47 nips2010-47-reference knowledge-graph by maker-knowledge-mining
Source: pdf
Author: Abhishek Kumar, Avishek Saha, Hal Daume
Abstract: This paper presents a co-regularization based approach to semi-supervised domain adaptation. Our proposed approach (EA++) builds on the notion of augmented space (introduced in E ASYA DAPT (EA) [1]) and harnesses unlabeled data in target domain to further assist the transfer of information from source to target. This semi-supervised approach to domain adaptation is extremely simple to implement and can be applied as a pre-processing step to any supervised learner. Our theoretical analysis (in terms of Rademacher complexity) of EA and EA++ show that the hypothesis class of EA++ has lower complexity (compared to EA) and hence results in tighter generalization bounds. Experimental results on sentiment analysis tasks reinforce our theoretical findings and demonstrate the efficacy of the proposed method when compared to EA as well as few other representative baseline approaches.
[1] Hal Daum´ III. Frustratingly easy domain adaptation. In ACL’07, pages 256–263, Prague, Czech Republic, June 2007. e
[2] Hal Daum´ III, Abhishek Kumar, and Avishek Saha. Frustratingly easy semi-supervised domain adaptation. In ACL 2010 e Workshop on Domain Adaptation for Natural Language Processing (DANLP), pages 53–59, Uppsala, Sweden, July 2010.
[3] Theodoros Evgeniou and Massimiliano Pontil. Regularized multitask learning. In KDD’04, pages 109–117, Seattle, WA, USA, August 2004.
[4] Mark Dredze, Alex Kulesza, and Koby Crammer. Multi-domain learning by confidence-weighted parameter combination. Machine Learning, 79(1-2):123–149, 2010.
[5] Andrew Arnold and William W. Cohen. Intra-document structural frequency features for semi-supervised domain adaptation. In CIKM’08, pages 1291–1300, Napa Valley, California, USA, October 2008.
[6] John Blitzer, Ryan Mcdonald, and Fernando Pereira. Domain adaptation with structural correspondence learning. In EMNLP’06, pages 120–128, Sydney, Australia, July 2006.
[7] Gokhan Tur. Co-adaptation: Adaptive co-training for semi-supervised learning. In ICASSP’09, pages 3721–3724, Taipei, Taiwan, April 2009.
[8] Wenyuan Dai, Gui-Rong Xue, Qiang Yang, and Yong Yu. Transferring Naive Bayes classifiers for text classification. In AAAI’07, pages 540–545, Vancouver, B.C., July 2007.
[9] Dikan Xing, Wenyuan Dai, Gui-Rong Xue, and Yong Yu. Bridged refinement for transfer learning. In PKDD’07, pages 324–335, Warsaw, Poland, September 2007.
[10] Lixin Duan, Ivor W. Tsang, Dong Xu, and Tat-Seng Chua. Domain adaptation from multiple sources via auxiliary classifiers. In ICML’09, pages 289–296, Montreal, Quebec, June 2009.
[11] Ming-Wei Chang, Michael Connor, and Dan Roth. The necessity of combining adaptation methods. In EMNLP’10, pages 767–777, Cambridge, MA, October 2010.
[12] Vikas Sindhwani, Partha Niyogi, and Mikhail Belkin. A co-regularization approach to semi-supervised learning with multiple views. In ICML Workshop on Learning with Multiple Views, pages 824–831, Bonn, Germany, August 2005.
[13] D. S. Rosenberg and P. L. Bartlett. The Rademacher complexity of co-regularized kernel classes. In AISTATS’07, pages 396–403, San Juan, Puerto Rico, March 2007.
[14] John Blitzer, Koby Crammer, Alex Kulesza, Fernando Pereira, and Jennifer Wortman. Learning bounds for domain adaptation. In NIPS’07, pages 129–136, Vancouver, B.C., December 2007.
[15] John Blitzer, Mark Dredze, and Fernando Pereira. Biographies, bollywood, boom-boxes and blenders: Domain adaptation for sentiment classification. In ACL’07, pages 440–447, Prague, Czech Republic, June 2007.
[16] Shai Ben-David, John Blitzer, Koby Crammer, and Fernando Pereira. Analysis of representations for domain adaptation. In NIPS’06, pages 137–144, Vancouver, B.C., December 2006.
[17] Piyush Rai, Avishek Saha, Hal Daum´ III, and Suresh Venkatasubramanian. Domain adaptation meets active learning. In e NAACL 2010 Workshop on Active Learning for NLP (ALNLP), pages 27–32, Los Angeles, USA, June 2010.
[18] Hal Daum´ III. Notes on CG and LM-BFGS optimization of logistic regression. August 2004. e
[19] Vikas Sindhwani and David S. Rosenberg. An RKHS for multi-view learning and manifold co-regularization. In ICML’08, pages 976–983, Helsinki, Finland, June 2008.
[20] Avrim Blum and Tom Mitchell. Combining labeled and unlabeled data with co-training. In COLT’98, pages 92–100, New York, NY, USA, July 1998. ACM.
[21] Maria-Florina Balcan and Avrim Blum. A PAC-style model for learning from labeled and unlabeled data. In COLT’05, pages 111–126, Bertinoro, Italy, June 2005.
[22] Maria-Florina Balcan and Avrim Blum. A discriminative model for semi-supervised learning. J. ACM, 57(3), 2010.
[23] Karthik Sridharan and Sham M. Kakade. An information theoretic framework for multi-view learning. In COLT’08, pages 403–414, Helsinki, Finland, June 2008. 9