Application of ANN for water quality index

Abstract

Attempt has been made to create a Water QualityIndex (WQI) based on artificial neural network (ANN) andglobally accepted parameters. Several methods to measureWQI are available in the research and ambiguity problemsexist where all the sub-indices of WQI are acceptable butoverall index is not acceptable. In this study, we have tried todevelop the WQI based on the WHO (world HealthOrganization) parameters (Dissolved Oxygen, pH, Turbidity, E.Coli and Electric Conductivity). The results also revealchanges in ANN based result from various input neuralnetwork model and its parameters. Evenwithin same model,changes occur with variation in parameter. Based on thestatistical parameter of regression value, theparameter andnetwork model would be selected. With the dataset created forthis study have shown the Cascade network is best forpredicting the WQI.