Abstract
For a classification result, traditional evaluation method is classification error. However, classification error might be not comprehensive for complex classifier evaluation. This week I will introduce some graph method for classification performance evaluation.
References
1. Prati, R. and Batista, G. and Monard, M., A survey on graphical methods for classification predictive performance evaluation. IEEE Trans. on Knowledge and Data Engineering 23(11), 1601-1618, 2011
PreviewTips
Download
To download attachments, please log in.
Last Modified: 2012/02/19 02:55 | Author: Dai Dai