Artificial neural networks based on principal component analysis, fuzzy systems and fuzzy neural networks for preliminary design of rubble mound breakwaters

BALAS C. E., Koc M. L., TÜR R.

APPLIED OCEAN RESEARCH, vol.32, no.4, pp.425-433, 2010 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 32 Issue: 4
  • Publication Date: 2010
  • Doi Number: 10.1016/j.apor.2010.09.005
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.425-433
  • Keywords: Artificial intelligence, Neural networks, Fuzzy sets, Rubble-mound breakwaters, RISK-ASSESSMENT, INPUT SELECTION, PREDICTION, REVETMENTS, LOGIC
  • Gazi University Affiliated: Yes


The new artificial intelligence models proposed for the preliminary design of rubble mound breakwaters consist of (1) multi layer feed forward artificial neural networks, (2) hybrid artificial neural networks with principal component analysis, (3) fuzzy systems, and (4) fuzzy neural networks. These models are applied for the stability analyses of Mersin yacht harbor main breakwater, as a case study in Turkey. A better agreement between the predicted stability numbers of hybrid artificial neural networks and measurements is obtained when compared to the stability equations. The Hybrid Artificial Neural Network model that is trained by the pre-processed database of measurements obtained from the Principal Component Analysis is considered as a robust technique in handling uncertainties inherent in the preliminary design. The fuzzy system and fuzzy neural network models have the advantages of incorporating flexible reasoning as expert systems when compared to hybrid neural networks; however, they require the development of new prediction enhancement techniques for the improvement of their forecasts. (C) 2010 Elsevier Ltd. All rights reserved.