JSE_V7_N3_RP1 Ghedhban Swadi Karim Al-jebory Dave Holifield Journal on Software Engineering 2230–7168 7 3 9 14 Simulation, Neural Network, Predictor, Sonar, Seabed In this paper, the performance of two dynamic neural network based predictors in seabed mapping is investigated. The two types of predictors are; the Focused Time- Delay Neural Network (FTDNN) based predictor and the Nonlinear Autoregressive Network with Exogenous Inputs (NARX) predictor. A testing platform has been developed that consists of seabed simulator and sonar simulator. Results show the NARX predictor outperforms the FTDNN predictor. January - March 2013 Copyright © 2013 i-manager publications. All rights reserved. i-manager Publications http://www.imanagerpublications.com/Article.aspx?ArticleId=2169