Beyond Independence: Conditions for the Optimality of the Simple Bayesian Classifier

@inproceedings{citeulike:2358549, abstract = {The simple Bayesian classifier (SBC) is commonly thought to assume that attributes are independent given the class, but this is apparently contradicted by the surprisingly good performance it exhibits in many domains that contain clear attribute dependences. No explanation for this has been proposed so far. In this paper we show that the SBC does not in fact assume attribute independence, and can be optimal even when this assumption is violated by a wide margin. The key to this finding lies in...}, author = {Domingos, Pedro and Pazzani, Michael J. }, booktitle = {International Conference on Machine Learning}, citeulike-article-id = {2358549}, keywords = {classification}, pages = {105--112}, posted-at = {2008-02-09 21:52:54}, priority = {2}, title = {Beyond Independence: Conditions for the Optimality of the Simple Bayesian Classifier}, url = {http://citeseer.ist.psu.edu/domingos96beyond.html}, year = {1996} }

TY - CONF ID - citeulike:2358549 L3 - citeulike-article-id:2358549 TI - Beyond Independence: Conditions for the Optimality of the Simple Bayesian Classifier T2 - International Conference on Machine Learning SP - 105 EP - 112 N2 - The simple Bayesian classifier (SBC) is commonly thought to assume that attributes are independent given the class, but this is apparently contradicted by the surprisingly good performance it exhibits in many domains that contain clear attribute dependences. No explanation for this has been proposed so far. In this paper we show that the SBC does not in fact assume attribute independence, and can be optimal even when this assumption is violated by a wide margin. The key to this finding lies in... KW - classification AU - Domingos, Pedro AU - Pazzani, Michael PY - 1996/// UR - http://citeseer.ist.psu.edu/domingos96beyond.html ER -