Cost-Sensitive Classifier Evaluation Using Cost Curves

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 outlines the most important requirements for cost-sensitive classifier evaluation, and introduces a technique for classifier performance visualization - the cost curve - that meets all these requirements. This talk should be of interest to anyone who works in areas that use classifiers, for example, machine learning, pattern recognition, biometrics, and diagnosis.

Speaker: Robert Holte, Univ. of Alberta