PREDICTION OF LIQUFICATION USING GROUP METHOD OF DATA HANDLING

Caissons are massive structures which are used for anchor or foundation of marine facilities. During large earthquakes, liquefaction is occurred and caissons may becompletely collapsed. Several researchers performed studies on liquefaction [1-5]. In recent years, artificial intelligent methods were applied to predict liquefaction.Rahman and Wang, (2002) used Fuzzy Logic for liquefaction predication. Jeng et al.,(2003) applied Artificial Neural Network to predict liquefaction[6,7] GMDH based on neural network neurons and these neurons makes it possible to predict complex system modeling. Recently, some researchers have been done about GMDHprediction in engineering problems. In this method in training period, weights are calculated by Least Square Estimation (LSE) method. Recently, some researches havebeen used GMDH for solving engineering problems [8-10] In this study, Group Method Handling was used to predict liquefaction in sand due to experimental database. They have compared GMDH correlation coefficient with genetic programming and neural network correlation coefficients. The results showed that GMDH is better than above mentioned methods.