Novel classifier scheme for imbalanced problems

There is an increasing interest in the design of classifiers for imbalanced problems due to their relevance in many fields, such as fraud detection and medical diagnosis. In this work we present a new classifier developed specially for imbalanced problems, where maximum F-measure instead of maximum accuracy guide the classifier design. Theoretical basis, algorithm description and real experiments are presented. The algorithm proposed shows suitability and a very good performance in imbalance scenarios and high overlapping between classes.