In this chapter, we consider the following problem of classification (prediction): for a decision table T and a new object v, given by values of conditional attributes from T, it is required to generate a decision corresponding to v. We compare qualities of classifiers based on exact deterministic and inhibitory decision rules. The first type of classifiers is the following: for a given decision table we construct for each row an exact deterministic decision rule using the greedy algorithm. The obtained system of rules jointly with simple procedure of voting can be considered as a classifier. A deterministic rule, which is realizable for given object, is a vote "pro" the decision from the right-hand side of the rule. © 2009 Springer-Verlag Berlin Heidelberg.
Bibliographical noteGenerated from Scopus record by KAUST IRTS on 2023-09-21
ASJC Scopus subject areas
- Artificial Intelligence