Designing fuzzy imbalanced classifier based on the subtractive clustering and Genetic Programming
Mahdizadeh Mahboubeh, Mahdi Eftekhari · 2013
In this paper, a design methodology is proposed for generating a fuzzy rule-based classifier for imbalanced datasets. The classifier is based on Sugeno-type Fuzzy Inference System. It is generated by using of subtractive clustering and Multi-Gene Genetic Programming to obtain fuzzy rules. The subtractive clustering is utilized for producing the antecedents of rules and Multi-Gene Genetic Programming is employed for generating the functions in the consequence parts of rules. Feature selection is utilized as an important pre-processing step for dimension reduction. Experiments are performed with 8 datasets from KEEL. The comparison results reveal that the proposed classifier outperforms the other methods.