Application of Artificial Neural Network Tools for the Prediction of Water Quality Index of Hand Dug Well Water in Gboko Local Government Area of Benue State

This research assessed the ability of artificial neural network tools to accurately predict water quality index. Water from hand dug wells within Gboko Local Government area of Benue state were sampled using standard procedures of the American Public Health Association (APHA, 2017). Twelve water quality indices, namely; pH, TDS, EC, Alkalinity, TH, BOD, Mg, Cl-, NO3, Fe and Zn were evaluated and their concentrations used to determine the Water quality index (WQI) of hand dug well water in Gboko Local Government Area, making use of the weighted arithmetic water quality index (WAWQI) method. The FeedForward Neural Network (FNN) Tool was designed using the MATLAB APP. The FNN tool was designed with twelve input parameters, 10 hidden neurons, and one output parameter (WQI). The performance validation of the ANN tool was achieved at the 8th run with a regression value greater than 0.8 and a mean square error of 0.6981.The FNN tool showed high accuracy in predicting the WQI. The accuracy of the results revealed that ANN tool can effectively be deployed in accurately predicting the WQI of a large dataset within a short time. Based on the findings, the study recommended the adaption of FNN tool for quicker and accurate determination of WQI.