Quality Measures for Semi-Automatic Learning of Simple Diagnostic Rule Bases Martin Atzmueller, Joachim Baumeister, Frank Puppe

Abstract. Semi-automatic data mining approaches often yield better
results than plain automatic methods, due to the early integration of
the user's goals. For example in the medical domain, experts are
likely to favor simpler models instead of more complex models.
Then, the accuracy of discovered patterns is often not the only
criterion to consider. Instead, the simplicity of the discovered
knowledge is of prime importance, since this directly relates to
the understandability and the interpretability of the learned knowledge.
In this paper, we present quality measures considering the
understandability and the accuracy of (learned) rule bases. We
describe an unifying quality measure, which can trade-off small
losses concerning accuracy vs.an increased simplicity. Furthermore,
we introduce a semi-automatic data mining method for learning
understandable and accurate rule bases.

The presented work is evaluated using cases from a real world
application in the medical domain.

Keywords. quality Measures, diagnostic scores, semi-automatic learning methods

BiBTeX:

@INPROCEEDINGS{ABP:04,
   author = {Martin Atzmueller, Joachim Baumeister, Frank Puppe},
   title = {{Quality Measures for Semi-Automatic Learning of Simple Diagnostic Rule Bases}},
   booktitle = {Proc. 15th International Conference of Declarative Programming and Knowledge Management (INAP-2004)},
   pages = {203-213},
   year = {2004}
}