A centerbulb was integrated to a catamaran in order to create a favorable secondary wave interference to reduce the unfavorable wave interference resistance between the demi-hulls. An Artificial Neural Network (ANN) technique is implemented together with a computational flow solver based on the low-Froude number theory, which uses a source-panel method with Dawson's algorithm, to find an efficient centerbulb geometry. In order to gain computation time in the optimization algorithm the present ANN approach is applied to the responses of the flow solver to the centerbulb design parameter changes. The numerical design procedure is verified by model tests conducted with a catamaran model both with and without the centerbulb. Numerical and experimental analyses verify each other and confirm that it is possible to reduce wave interference in considerable amounts with an optimal centerbulb. The study presented for catamarans, demonstrates the capabilities of the optimization procedure in reducing the wave interference resistance as well as the efficiency and ease of application of the centerbulb as a retrofit. (C) 2014 Elsevier Ltd. All rights reserved.