Machine Learning Approach for Predicting the Quality of Water

Abstract

The problem of water quality is increasing with every passing day due to number of causes like urbanization, waste material from industries, rapid growth in population, sewage disposal, radioactive waste, plastics and other harmful chemicals. Very often our own actions leads to that pollution. The need of the hour is to solve this problem and get the good water for various purposes. Our work mainly focuses on building a machine learning model for water quality analysis of Indian rivers. We classify the dataset into four classes using the water quality standards provided by the central pollution control board of India. Total of eight   top features after feature engineering are considered in building that classification model like Dissolved oxygen, temperature, Biochemical oxygen dissolve, pH value etc. Our machine learning model gives the accuracy of 96.1%, which was better than all the previous work done on predicting the quality of water.