SVM Model to Predict the Water Quality Based on Physicochemical Parameters

Abstract

Analysis of water quality is a very important and challenging task in the management of water bodies and requires immediate attention as it adversely affects the health of living beings. Three parameters namely, pH, Total Dissolved Solids (TDS), and Turbidity were used for data analysis. In this study for mapping of training samples from input space to higher dimensional feature space, LibSVM, (a library of SVM) was used with the use of two kernel function types Radial Basis Function and Polynomial function. For performing the experiment, the three parameter combinations (C, d, ϒ) were evaluated based upon the kernel by taking various range values to obtain the best type of kernel functions through a 10-fold cross-validation process. After performing all experiments, a comparative analysis was done to evaluate the best parameter combination (C, d, ϒ) and the values of performance measures. The result shows that the optimum model developed using LibSVM with the use of Polynomial Kernel function which gives an accuracy of 99.434% in predicting water quality.

Keywords

• support vector machine
• water quality
• libsvm
• kernel functions
• nainital lake