Neuro-fuzzy classification of the Rhagoletis pomonella species group using digitized wing structures

In this paper, we applied a neuro-fuzzy system to classify the morphologically indistinguishable Rhagoletis pomonella sibling species group. A carefully selected set of wing structure and shape variables were fuzzified using triangular membership functions. The neuro-fuzzy system NEFCLASS was applied to train the fly morphological datasets and a set of fuzzy rules were constructed. A fuzzy inference engine was constructed using the fuzzy rule bases. Furthermore, manually pruned fuzzy rules were employed to make a fuzzy key to classify this sibling species group.