Water Quantity Prediction Using Least Squares Support Vector Machines (LS-SVM) Method

Nian Zhang, Charles Williams, Pradeep Behera

The impact of reliable estimation of stream flows at highly urbanized areas and the associated receiving waters is very important for water resources analysis and design. We used the least squares support vector machine (LS-SVM) based algorithm to forecast the future streamflow discharge. A Gaussian Radial Basis Function (RBF) kernel framework was built on the data set to optimize the tuning parameters and to obtain the moderated output. The training process of LS-SVM was designed to select both kernel parameters and regularization constants. The USGS real-time water data were used as time series input. 50% of the data were used for training, and 50% were used for testing. The experimental results showed that the LS-SVM algorithm is a reliable and efficient method for streamflow prediction, which has an important impact to the water resource management field.

Full Text