A novel rules optimizer with feature selection using rough-entropy-coverage partitioning based reduci

This paper presents a novel approach for optimizing the number of decision rules and select important features based on reduct. We compute the reduct using entropy value of conditional attribute then eradicates the redundant dataset, noisy features and uncertainty of dataset using coverage factor and generate optimized number of rules. Experimental results show that this approach achieves high data reduction with important feature selection as well as optimize the number of rules compared to earlier works.