Predicting the parameters of water quality and calculating the Water Quality Index of Ulsoor Lake, Bangalore, India using Deep Learning Techniques

Water is one of the precious natural resources present on earth. Water quality prediction is an important challenge for researchers. The increasing intervention of human activities has led to the deterioration of water quality of different water bodies, causing threat to the aquatic creatures. This paper aims at predicting the values of different parameters that are used to determine the water quality of the lake by comparing seven algorithms namely BPNN, SVR, SRN, LSTM, GRU, Bi-LSTM and Improved RNN, calculating the Water Quality Index using weighted arithmetic water quality index (W A WQI) method and suggesting methodologies to improve the water quality. The models are evaluated using 3 evaluation metrices namely RMSE, MAE and MAPE. The data from Ulsoor lake, Bengaluru, Karnataka, India is used to verify the proposed model.