%0 Journal Article %T Simulation and Prediction of Groundwater Quality of a Semi-Arid Region Using Fuzzy Inference System and Neural Network Techniques %J Journal of Soft Computing in Civil Engineering %I Pouyan Press %Z 2588-2872 %A Palanichamy, Jegathambal %A Palani, Sundarambal %A Hebsiba, G. Anita %A Viola, Jansi %A Tungsrimvong, Apinun %A Babu, Babithesh %D 2022 %\ 01/01/2022 %V 6 %N 1 %P 110-126 %! Simulation and Prediction of Groundwater Quality of a Semi-Arid Region Using Fuzzy Inference System and Neural Network Techniques %K Artificial Neural Network %K Fuzzy logic %K Groundwater quality %K Supervised machine learning %K WQI %K Soft computing techniques %R 10.22115/scce.2022.285106.1314 %X The groundwater is the main source of domestic and agricultural purposes in the arid and semi-arid regions where the surface water availability is limited. To protect and manage the groundwater system effectively, a thorough knowledge and understanding of groundwater quality and application of computational methods to simulate the complex and nonlinear groundwater system are paramount necessary. Generally, three types of models such as physically based model, conceptual models and Blackbox models are applied to study the interconnected processes in the subsurface media. In this study, Artificial Neural Network (ANN) (3 Models with 1, 2 and 3 outputs) was used to simulate and predict the concentration of groundwater quality parameters and Mamdani Fuzzy Inference System (MFIS) was used to simulate the water quality indices. Classification algorithms of NEUROSHELL and MATLAB were used to predict the class of items in a data set. The model was constructed using already-labelled items of similar data sets. The WQI of 29 samples was determined using weighted average method. Based on MFIS, 10 samples were classified as ‘good’, four samples as ‘poor’ and remaining samples as ‘very poor’. The simulation model using the classification algorithm of ANN was used to predict the concentration of groundwater quality parameters and it was observed that three ANN models values and the actual data fit well with correlation coefficient varying from 0.93 to 0.99. When the soft computing techniques can be coupled with geospatial and geostatical method to map the spatial and temporal distribution of water quality parameters. %U https://www.jsoftcivil.com/article_145347_f687a69ad20df99aef4db744e4e14b65.pdf