Čače, IvanaMeyer, John-Jules Ch.Pieterman, C.R.C.Valk, G.D.Horbach, Matthias2019-03-072019-03-072013978-3-88579-614-5https://dl.gi.de/handle/20.500.12116/20614In this paper we show how knowledge from the medical domain can be incorporated in classification in a way that improves the transparency of classification (the 'why'), and makes the classification less dependent on both the particular data set used for training and on peculiarities of the classification algorithm. We compare a decision tree incorporating a domain model with a tree built directly from the same data.enText/Conference Paper1617-5468