| Building intelligent systems in open, heterogeneous, dynamic, distributed environments | 16 May 2008, 08:30:08 EDT ![]() |
|||
Cost-Sensitive Classifier Evaluation Using Cost Curves Title: Cost-Sensitive Classifier Evaluation Using Cost Curves Speaker: Robert Holte Start Date: Thursday, September 28, 2006, 10:00AM End Date: Thursday, September 28, 2006, 11:00AM Location: 325b Abstract: The evaluation of classifier performance in a cost-sensitive setting is straightforward if the operating conditions (misclassification costs and class distributions) are fixed and known. When this is not the case, evaluation requires a method of visualizing classifier performance across the full range of possible operating conditions. This talk argues that the classic technique for classifier performance visualization -- the ROC curve – is inadequate for the needs of researchers and practitioners in several important respects. It then describes a different way of visualizing classifier performance -- the cost curve -- that overcomes these deficiencies. No familiarity with ROC curves or cost curves is necessary, they will be fully explained. Tags: learning, visualization, hci Host: Marie desJardins |
| Home | About Us | Contact Us | Site Map | Legal | Privacy Copyright © 1999-2008 UMBC ebiquity research group. Copyright © 2003-2008 Site design and RGB engine code by Filip Perich. XG Page gen 0.027 sec. |