Wing Pattern-Based Classification of the Rhagoletis pomonella Species Complex Using Genetic Neural Networks

The Rhagoletis pomonella species complex consists of at least four sibling species. They are highly host specific as larvae, and display great fidelity as adults. The only certain way to identify them is to know the host materials from which they came, because these fruit flies are very similar or identical, and have been especially recalcitrant to morphological separation. In this paper we hypothesize that there is hidden biological information in the wing vein structure in the pomonella species group that can be used to distinguish them. Classification of the species complex is modeled via Bayesian and probability neural networks using information on wing size, shape and vein structure. The classification models were optimized through a genetic algorithm by selecting the optimal features and performed well in classifying new specimens. The results have implications for agricultural production and quarantine issues and could be helpful in devising a classification system for rapid id...

• Type: Journal
• Year: 2007
• Where: IJCSA
• Authors: Chengpeng Bi, Michael C. Saunders, Bruce A. McPheron
• Keywords: IJCSA 2007, Rhagoletis Pomonella, Vein Structure, Wing Vein Structure