Analysis of Support Vector Machine Model in Forecasting the One Year Ahead Water Quality of the River Ganga

Document Type : Primary Research paper


1 Computer Science & Information Technology, Mahatma Jyotibha Phule Rohilkhand University, Bareilly, U.P. India

2 Deptt. Of Environment Science, Gurukul Kangri University, Haridwar, Uttarakhand, India

3 Deptt. Of Computer Science, BIAS, Bhimtal,Uttarakhand, India


Analysis of river water quality forecasting model based on Support Vector Machine (SVM) for the Ganges River has not been reported in the literature to the best knowledge of the authors. This paper attempts to develop such models for the Ganges River in the area from Devprayag to Roorkee, Uttarakhand, India. A monthly experimental dataset has been used from 2001 to 2015 time series. This paper examines the possibilities of SVM techniques for the 2016 water quality forecast of the Ganges River based on a continuous dataset of the last fifteen years. In conclusion, the SVM-based technique failed to develop a year ahead forecasting models (as compared to Artificial Neural Network (ANN) based previous work) that could efficiently estimate the water quality of the Ganges River. In the future, some other modifications will be made in the proposed model in order to have efficient forecasting.