A Nguyen Thi Thu Nga creator Luketina, D.A. Chairperson Mark, Ole Examination committee Sutat Weesakul Examination committee Tawatchai Tingsanchali Examination committee DANIDA Scholarship donor text series technical report th Bangkok Asian Institute of Technology 2002 continuing eng 99, [28] leaves Salinity intrusion plays an important role in human being, including the effects to the water supply for agriculture and domestic. Although this field of study drew the attention of scientist from a long time ago, using Artificial Neural Networks (ANNs) for prediction salinity distribution in the estuary is still rarely been done. Two neural network software packages (EasyNN and WinNN32) are presented in this study for forecasting salinity twelve hours ahead at six potential water abstraction locations along the West Pearl River Estuary. The data used for training and testing neural networks by these packages came from the results of a previous study using the three-dimensional hydrodynamic model RMA. The results for different types of ANNs applied by EasyNN and WinNN32 are compared. The WinNN32 with quick-prop algorithm was found more suitable for the salinity prediction than the EasyNN. The results also indicate that with appropriate structures, ANNs can give high performance for salinity prediction in estuary. A thesis report submitted in partial fulfillment of the requirements for the degree of Master of Engineering Thesis (M.Eng.) - Asian Institute of Technology, 2002 Neural networks (Computer science) Salinity China West Pearl River Asian Institute of Technology. http://203.159.5.9/ait-thesis/detail.php?q=B07769 http://203.159.5.9/ait-thesis/detail.php?q=B07769 020823 20260818165857.0