Analyzing the Effectiveness and Applicability of Co-training

@inproceedings{citeulike:142759, abstract = {Recently there has been significant interest in supervised learning algorithms that combine labeled and unlabeled data for text learning tasks. The co-training setting [1] applies to datasets that have a natural separation of their features into two disjoint sets. We demonstrate that when learning from labeled and unlabeled data, algorithms explicitly leveraging a natural independent split of the features outperform algorithms that do not. When a natural split does not exist, co-training...}, author = {Nigam, Kamal and Ghani, Rayid }, booktitle = {CIKM}, citeulike-article-id = {142759}, keywords = {cotraining, probabilistic-models}, pages = {86--93}, posted-at = {2008-01-23 16:45:13}, priority = {3}, title = {Analyzing the Effectiveness and Applicability of Co-training}, url = {http://citeseer.ist.psu.edu/nigam00analyzing.html}, year = {2000} }

TY - CONF ID - citeulike:142759 L3 - citeulike-article-id:142759 TI - Analyzing the Effectiveness and Applicability of Co-training T2 - CIKM SP - 86 EP - 93 N2 - Recently there has been significant interest in supervised learning algorithms that combine labeled and unlabeled data for text learning tasks. The co-training setting [1] applies to datasets that have a natural separation of their features into two disjoint sets. We demonstrate that when learning from labeled and unlabeled data, algorithms explicitly leveraging a natural independent split of the features outperform algorithms that do not. When a natural split does not exist, co-training... KW - cotraining KW - probabilistic-models AU - Nigam, Kamal AU - Ghani, Rayid PY - 2000/// UR - http://citeseer.ist.psu.edu/nigam00analyzing.html ER -