In this paper, a new constructive approach for the automatic definition of feedforward neural networks (FNNs) is introduced. Such approach (named MASCoNN) is multiagent-oriented and, thus, can be regarded as a kind of hybrid (synergetic) system. MASCoNN centers upon the employment of a two-level hierarchy of agent-based elements for the progressive allocation of neuronal building blocks. By this means, an FNN can be considered as an architectural organization of reactive neural agents, orchestrated by deliberative coordination entities via synaptic interactions. MASCoNN was successfully applied to implement nonlinear dynamic systems identification devices and some comparative results, involving alternative proposals, are analyzed here. © Springer-Verlag Berlin Heidelberg 2003.2810 462473Dahmen, W., Micchelli, C.A., Some remarks on ridge functions (1987) Approximation Theory and Its Applications, 3 (2-3), pp. 139-143Davis, P.J., (1975) Interpolation & Approximation, , Dover Publications, New YorkFahlman, S.E., Lebiere, C., The cascade-correlation learning architecture (1990) Advances in Neural Information Processing Systems, 2, pp. 524-532. , D. S. Touretzky, editor, Morgan KaufmannGhosh, J., Neural-symbolic hybrid systems (2001) The Handbook of Applied Computational Intelligence, , M. Padget et al., editors, CRC Press(1995) Intelligent Hybrid Systems, , S. Goonatilake and S. Khebbal, editors. WileyHärdle, W., (1990) Applied Nonparametric Regression, , Cambridge University PressHaykin, S., (1999) Neural Networks-A Comprehensive Foundation, , Prentice HallHwang, J., Lay, S., Maechler, M., Martin, D., Schimert, J., Regression modeling in backpropagation and project pursuit learning (1994) IEEE Trans. on Neural Networks, 5 (3), pp. 342-353. , MayKwok, T.-Y., Yeung, D.-Y., Constructive algorithms for structure learning in feedforward neural networks for regression problems (1997) IEEE Trans. on Neural Networks, 8 (3), pp. 630-645Medskar, L.A., (1995) Hybrid Intelligent Systems, , Kluwer Academic PublisherNarendra, K., Parthasarathy, K., Identification and control of dynamical systems using neural networks (1990) IEEE Trans. on Neural Networks, 1 (1), pp. 4-27. , MarchNrgaard, M., Ravn, O., Poulsen, N.K., Norgaard, P.M., Hansen, L.K., Neural Networks for Modelling and Control of Dynamic Systems: A Practitioner's Handbook (2000) Advanced Textbooks in Control and Signal Processing, , SpringerPham, K.M., The NeurOagent: A Neural Multi-agent Approach for Modelling, Distributed Processing and Learning, pp. 221-244. , Goonatilake and Khebbal [5], chapter 12Reed, R., Pruning algorithms-A survey (1993) IEEE Trans. on Neural Networks, 4 (5), pp. 740-747. , MayScherer, A., Schlageter, G., A Multi-agent Approach for the Integration of Neural Networks and Expert Systems, pp. 153-173. , Goonatilake and Khebbal [5], chapter 9Selfridge, O.G., Pandemonium: A paradigm for learning (1958) Proc. Symp. Held Physical Lab.: Mechanisation Thought Processing, pp. 511-517. , LondonŚmieja, F.J., The pandemonium system of reflective agents (1996) IEEE Trans. on Neural Networks, 7 (1), pp. 97-106. , JanuaryTaha, I., Ghosh, J., Symbolic interpretation of artificial neural networks (1999) IEEE Trans. on Knowledge and Data Eng., 11 (3), pp. 448-463. , May/JuneZuben, F.J.V., Netto, M., Projection pursuit and the solvability condition applied to constructive learning (1997) Proc. of the International Joint Conference on Neural Networks, 2, pp. 1062-1067