Cost-Sensitive Learning with Neural Networks.

In the usual setting of Machine Learning, classifiers are typically evaluated by estimating their error rate (or equi valently, the classification accuracy) on the test data. However, this mak es sense only if all errors have equal (uniform) costs. When the costs of er- rors differ between each other, the classifiers should be eva luated by comparing the total costs of the errors. Classifiers are typically designed to minimize the number of errors (incorrect classifications) made. When misclassification c osts vary between classes, this approach is not suitable. In this case the total misclassification cost should be minimized. In Machine Learning, only little work for dealing with non- uniform misclassification costs has been done. This paper pr esents a few different approaches for cost-sensitive modifications of the back- propagation learning algorithm for multilayered feedforw ard neural networks . The described approaches are thoroughly tested and eval- uated on several standard benchmark domains.