OCEAn: Ordinal classification with an ensemble approach

Generally, classification problems catalog instances according to their target variable without considering the relation among different labels. However, there are real in which values of class related each other. Because interest this type problem, several solutions have been proposed, such as cost-sensitive classifiers. Ensembles proven be very effective for tasks; however, far we know, no proposals that use a genetic-based methodology metaheuristic create models. In paper, present OCEAn, an ordinal algorithm based on ensemble approach, makes final prediction weighted vote system. This voting takes into account weights obtained by genetic tries minimize cost classification. To test performance compared our proposal with algorithms literature and demonstrated that, indeed, approach improves previous results.