Water Quality Prediction for Hanjiang with Optimized Support Vector Regression

Water quality prediction is of great significance to the policies of water quality control which are made by the government. In this paper, the prediction problem of Hanjiang water quality is investigated with genetic algorithm (GA) and support vector regression (SVR). Based on the collected water quality data of Hanjiang, SVR is utilized to train the prediction model, and achieve the result of water quality prediction. In order to obtain higher prediction accuracy, particle swarm optimization algorithm (PSO), grid search (GS) and GA are chosen to optimize the parameters of SVR, respectively. Experimental results show that the prediction accuracy of GA-SVR algorithm is significantly higher than SVR, PSO-SVR and GS-SVR.