Summary
International Technical Conference on Circuits/Systems, Computers and Communications
2016
Session Number:W2-1
Session:
Number:W2-1-4
Comparison of ANN and SVM for Prediction of Biochemical Oxygen Demand in Chaophraya River
Weeris Treeratanajaru, Supawin Watcharamul, Rajalida Lipikorn ,
pp.791-793
Publication Date:2016/7/10
Online ISSN:2188-5079
Summary:
Artificial Neural Network (ANN) and Support Vector Machine (SVM) models are used increasingly to predict, monitor and forecast water quality. In this paper, two methods were implemented to predict biochemical oxygen demand (BOD) of Chaophraya River, Thailand using a set of simple measurable surface water quality variables including water temperature, dissolved oxygen (DO), electrical conductivity (EC), pH, nitrate, ammonia, total phosphate (TP), monitoring time, and monitoring location as input variables. The data set consists of 1248 water samples represent 18 different monitoring stations along the Chaophaya, which has been monitored for 17 years. The associated parameters for optimum ANN and SVM model were obtained using grid search technique. The ANN and SVM models can predict BOD in training and testing data sets with reasonably high correlation. The overall results showed that both models could be used as one of the fast, reliable and cost-effective methods for predicting BOD in environments.