Identyfikatory
Warianty tytułu
Języki publikacji
Abstrakty
Paper deals with the knowledge acquisition process. Different experts formulate the rules for decision support systems. We assume they have different knowledge about the problem and therefore obtained rules have different qualities. We will formulate the proposition of the confidence measure and its application to the decision process. We will propose how calculate the value of measure under consideration for typical statistical learning process. On the base on the proposed measure of the knowledge quality we propose the procedure of the contradictions elimination for the set of logical rules.
Rocznik
Tom
Strony
MI115--119
Opis fizyczny
Bibliogr. 8 poz.
Twórcy
autor
- Chair of Systems and Computer Networks, Faculty of Electronics, Wroclaw University of Technology. Wybrzeze Wyspianskiego 27, 50-370 Wroclaw, Poland
Bibliografia
- [1] BERGADANO F., MATWIN S., MICHALSKI R.S., Zhang J., Measuring of Quality Concept Descriptions, Proc. Of the 3rd European Working Session on Learning, Aberdeen, Scotland, 1988.
- [2] BRUHA I., KOCKOVA S., Quality of decision rules: Empirical and statistical approaches, Informatica, no 17.
- [3] DEAN P., FAMILI A., Comparative Performance of Rule Quality Measures in an Inductive Systems, Applied Intelligence, no 7, 1997.
- [4] GIAKOUMAKIS E., PAPAKONSTANTIOU G., SKORDALAKIS E., Rule-based systems and pattern recognition, Pattern Recognition Letters, No 5, 1987.
- [5] GUR-ALI O., WALLANCE W.A., Induction of rules subject to a quality constraint: probabilistic inductive learning, IEEE Transaction on Knowledge and Data Engeineering, vol. 5, no 3, 1993.
- [6] MITCHELL T., Machine Learning, McGraw Hill, 1997.
- [7] SACHS L., Applied Statistic. A Handbook of Techniques, Springer-Verlag, New York Berlin Heideberg Tokyo, 1984.
- [8] WOZNIAK M., Idea of Knowledge Acquisition for the Probabilistic Expert Systems, Proceedings of the International Conference on Computational Intelligence for Modeling, Control and Automation, Vienna (Austria), 12-14 February 2003.
Typ dokumentu
Bibliografia
Identyfikator YADDA
bwmeta1.element.baztech-article-PWA4-0019-0017