Deterministic and nondeterministic decision rules in classification process
Piotr Paszek, Barbara Marszaë-Paszek · Journal of Medical Informatics & Technologies · 2010
In this paper an algorithm of calculating nondeterm inistic decision rules from the decision table was presented. The algorithm uses additional conditions imposed on rules. This is a greedy algorithm. The nondetermin istic decision rules were used in the process of classification of new examples, for medical data sets. The decision tables from the UCI Machine Learning Repository were used. The achieved results allow us to state that nondeterministi c decision rules can be used for improving the quality of classifica tion.