Malgorzata Latuszynska, Ph., D., Institute of IT in Management, University of Szczecin, Mickiewicza 64, Szczecin, 71 101, Poland and Mariusz Borawski, Ph., D., Faculty of Information Science, Technical University of Szczecin, Zolnierska 49, Szczecin, 71-211, Poland.
Data describing quantitative attributes of objects are often strongly dependent on each other. In some cases this dependence is strong to such the extent, that the data can not be considered separately. Then the data should be joined together to one abstract object. For this object arithmetic operations are defined. The example of such strongly dependent data are distribution and a variance.
Generalizing the notion of a distribution and a variance we can create a numerical set with elements governed by laws of both algebra and statistics during arithmetic operation execution. The paper will describe calculus for abstract objects based on a numerical set and show examples of applications in simulation models.