Using genetic algorithms to optimize the number of classification rules in SUCRAGE

Summary form only given. SUCRAGE is a supervised learning method by automatic generation of classification rules. The obtained results in generalization using the built rules are satisfactory. However, to be easily interpreted and to allow the explanation of the obtained classification, the rule base size must be reasonable. We propose to optimize the number of rules generated by SUCRAGE using genetic algorithms. The rule selection problem is formulated as a combinatorial optimization problem with two objectives: to maximize the number of correctly classified patterns and to minimize the number of classification rules. A set of rules is coded into a binary string and treated as an individual in genetic algorithm. A computer implementation of this optimization method is proposed and the experimental results obtained on various data are presented. The good performance of this approach allows us to make of SUCRAGE a knowledge acquisition tool, and to envisage the tests' extension to other data types.