Comparison of two prototype-based multicriteria classification methods

In the present study, we introduce an experimental analysis conducted over two Multicriteria Decision Aid (MCDA) classification methods that have been successfully applied to real world problems. Different from other studies on MCDA classifiers, which put more emphasis on the development of new methods, this work compares the effectiveness of two MCDA nominal classifiers under the same conditions. In such regard, we developed a customized genetic algorithm to calibrate their control parameters automatically under some different sets of reference alternatives, also known as prototypes. Moreover, the experiments we have realized so far, involving different datasets, reveal that there are still some gaps in existing MCDA methods that can lead to an improvement of the methodology for the classification problems. We understand that this sort of empirical assessment is interesting as it reveals how robust/sensitive the MCDA classifiers could be to the choice of cut-off thresholds and prototypes for distinct problems.