Prediction Of Creek Water Quality Based On Least Squares Vector Machine

Recent years, with the rapid development of the society economic, the pollution of water environmental is getting worse. In order to ensure the sustainable development of water resources in China. Strengthening the management of water resources and protecting water resources are one of the major issues which China’s water environment workers need to solve. The prediction of water quality is one primary means of water resources management and water pollution control, the indispensable basic work of water governance and water resources development and utilization, and the basic work of successful planning and management of the water environment and water pollution prevention and control tasks.Support vector machine is a new algorithm based on statistical learning theory which is developed recently. It’s a forefront research disciplines in the field of complex non-linear and artificial intelligence. Due to its outstanding classification and regression performance. Gradually it has a wide range of applications and research in many fields. This article attempts to introduce the support vector machine into the creek water quality prediction analysis. Do some exploratory research that applying least squares support vector machine to the creek water quality forecasts. In order to be able to find more accurate water quality prediction model than traditional forecasting methods.This paper first describes some mature and commonly used creek water quality model. And then analyzes several commonly used creek water quality prediction methods, which focused on the BP neural network algorithm and modeling steps. As well as the characteristics of the commonly used method. Theoretical basis and the principle of support vector machine learning is the focus of this article, including the content of the theory of machine learning, statistical learning theory and structural risk minimization principle. In this paper a more detailed study about a improvement form of the standard support vector machine-least squares support vector machine (LS-SVM) algorithm. This paper introduces LS-SVM modeling parameter selection and the choice of kernel function. On the basis of the above theoretical study. The emphasis is on using the least squares support vector machine to predict the future of the creek water quality. And a comparative analysis and BP neural network algorithm. Experiments show that:LS-SVM shows the superiority of the creek water quality prediction. It proves that the support vector machine approach to the introduction of the creek water quality prediction analysis is feasible and effective. It also shows that the outstanding theoretical advantages of the predictive ability of the support vector machine.