C4.5: programs for machine learning
C4.5: programs for machine learning
Machine Learning
Data Analysis and Mining in Ordered Information Tables
ICDM '01 Proceedings of the 2001 IEEE International Conference on Data Mining
Hi-index | 0.00 |
This article proposes alternative classification rules for the clubs in the Brazilian Soccer Championship employing Rough Sets Theory. A feature of the procedure adopted to determine the rough sets is that it allows for attributes associated to intransitive relations. One of such attributes is the result of direct confrontation. The classification procedure follows three stages. First, the condition attributes are ranked according to their importance for the official classification on points. Then the condition attributes are sequentially applied to generate a classification. Finally the result of this sequential classification is compared to the decision partition. The results obtained for different sets of attributes and different years are consistent, with the number of distinct classes obtained increasing with the quality of approximation.