Diverse Classes of Interval-Valued Aggregation Functions in Medical Diagnosis Support

In this contribution results connected with using new types of aggregation functions in medical diagnosis support are presented. These aggregation functions belong to the recently introduced families of possible and necessary aggregation functions as well as aggregation functions with respect to admissible linear orders. Examples of the mentioned families of aggregation functions proved to be comparably effective (if it comes to statistical measures and lower cost of prediction) to the previously used aggregation functions in medical diagnosis support systems. The considered classes of aggregation functions differ from the ones previously applied by the comparability relations between intervals involved in the monotonicity conditions.