%0 Conference Paper %1 YangWebb03 %A Yang, Y. %A Webb, G.I. %B Lecture Notes in Artificial Intelligence Vol. 2637: Proceedings of the Seventh Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD'03) %C Berlin/Heidelberg %D 2003 %E Whang, K-Y. %E Jeon, J. %E Shim, K. %E Srivastava, J. %I Springer-Verlag %K Bayes Discretization Naive for %P 501-512 %T Weighted Proportional k-Interval Discretization for Naive-Bayes Classifiers %X The use of different discretization techniques can be expected to affect the bias and variance of a learning algorithm. We call such an effect discretization bias and variance. Proportional k-interval discretization (PKID) tunes discretization bias and variance by adjusting discretized interval size and number proportional to the number of training instances. Theoretical analysis suggests that this is desirable for naive-Bayes classifiers. However PKID has sub-optimal performance when learning from small training data. We argue that this is because PKID equally weighs bias reduction and variance reduction. But for small data, variance reduction can contribute more to lower learning error and thus should be given greater weight than bias reduction. Accordingly we propose weighted proportional k-interval discretization (WPKID), which establishes a more suitable bias and variance trade-off for small data while allowing additional training data to be used to reduce both bias and variance. Our experiments demonstrate that for naive-Bayes classifiers, WPKID improves upon PKID for smaller datasets with significant frequency; and WPKID delivers lower classification error significantly more often than not in comparison to the other three leading alternative discretization techniques studied.

@inproceedings{YangWebb03, abstract = {The use of different discretization techniques can be expected to affect the bias and variance of a learning algorithm. We call such an effect discretization bias and variance. Proportional k-interval discretization (PKID) tunes discretization bias and variance by adjusting discretized interval size and number proportional to the number of training instances. Theoretical analysis suggests that this is desirable for naive-Bayes classifiers. However PKID has sub-optimal performance when learning from small training data. We argue that this is because PKID equally weighs bias reduction and variance reduction. But for small data, variance reduction can contribute more to lower learning error and thus should be given greater weight than bias reduction. Accordingly we propose weighted proportional k-interval discretization (WPKID), which establishes a more suitable bias and variance trade-off for small data while allowing additional training data to be used to reduce both bias and variance. Our experiments demonstrate that for naive-Bayes classifiers, WPKID improves upon PKID for smaller datasets with significant frequency; and WPKID delivers lower classification error significantly more often than not in comparison to the other three leading alternative discretization techniques studied.}, added-at = {2016-03-20T05:42:04.000+0100}, address = {Berlin/Heidelberg}, audit-trail = {Waiting on copy of copyright form from Ying}, author = {Yang, Y. and Webb, G.I.}, biburl = {https://www.bibsonomy.org/bibtex/2bca4c6b308058294fa599ebbedb4d84a/giwebb}, booktitle = {Lecture Notes in Artificial Intelligence Vol. 2637: Proceedings of the Seventh {Pacific}-{Asia} Conference on Knowledge Discovery and Data Mining (PAKDD'03)}, editor = {Whang, K-Y. and Jeon, J. and Shim, K. and Srivastava, J.}, interhash = {d1bc979a5adf52efaaa2cf739bc8db80}, intrahash = {bca4c6b308058294fa599ebbedb4d84a}, keywords = {Bayes Discretization Naive for}, location = {Seoul, Korea}, pages = {501-512}, publisher = {Springer-Verlag}, timestamp = {2016-03-20T05:42:04.000+0100}, title = {Weighted Proportional k-Interval Discretization for Naive-Bayes Classifiers}, year = 2003 }