Bianconi, André [UNESP]Zuben, Claudio Jose Von [UNESP]Serapião, Adriane Beatriz De Souza [UNESP]Govone, José Silvio [UNESP]2016-04-012016-04-012010Australian Journal of Entomology, v. 49, p. 201-212, 2010.1326-6756http://hdl.handle.net/11449/137319Artificial neural networks (ANNs) have been widely applied to the resolution of complex biological problems. An important feature of neural models is that their implementation is not precluded by the theoretical distribution shape of the data used. Frequently, the performance of ANNs over linear or non-linear regression-based statistical methods is deemed to be significantly superior if suitable sample sizes are provided, especially in multidimensional and non-linear processes. The current work was aimed at utilising three well-known neural network methods in order to evaluate whether these models would be able to provide more accurate outcomes in relation to a conventional regression method in pupal weight predictions of Chrysomya megacephala, a species of blowfly (Diptera: Calliphoridae), using larval density (i.e. the initial number of larvae), amount of available food and pupal size as input data. It was possible to notice that the neural networks yielded more accurate performances in comparison with the statistical model (multiple regression). Assessing the three types of networks utilised (Multi-layer Perceptron, Radial Basis Function and Generalised Regression Neural Network), no considerable differences between these models were detected. The superiority of these neural models over a classical statistical method represents an important fact, because more accurate models may clarify several intricate aspects concerning the nutritional ecology of blowflies.201-212engBlowflyLarval densityMass rearingNeural algorithmPupal weightArtigoAcesso restrito534874933112053699781434318986075628510167953810000-0001-9728-70920000-0002-9622-3254