H. Wu and J.M. Mendel (USA)
type-2 fuzzy logic system, classifier, ground vehicle,acoustic, majority voting
In this paper we present our preliminary investigation for the multiple-category classification of ground vehicles based on their acoustic emissions. By analyzing the fea tures within each run and across runs, we found that the run-means and run-standard-deviations of the features vary from run to run for all kinds of vehicles, and we there fore used type-2 fuzzy sets to model the uncertainties con tained in these features and constructed a type-2 fuzzy logic rule-based classifier (FL-RBC). To evaluate the type 2 FL-RBC in a fair way, we also constructed the Bayesian classifier and type-1 FL-RBC, and compared their perfor mances through leave-one-out experiments. Our experi ments showed that the type-2 FL-RBC has the best perfor mance among all three classifiers, and the adaptive process ing of block decisions based on the majority voting strategy can greatly improve the classifier's performance.
Important Links:
Go Back
