Water Quality Analysis and Prediction using Machine Learning Algorithms

The main objective of this work is to measure water quality using machine learning algorithms. A Water Quality Index (WQI) is a numeric expression used to evaluate the quality of a given water body. In this paper the following water quality parameters were used to evaluate the overall water quality in terms of the WQI. These parameters were as temp, dissolved oxygen (DO) (% sat), pH, conductivity, Biochemical oxygen demand (BOD), nitrates (NO3), faecal and total coli forms (TC). These parameters are used as feature vector to represent the water quality. In paper five kinds of classification algorithms, namely Navie Bayes(NB), Decision Tree(DT), K-Nearest Neighbor (KNN), Support Vector Machine(SVM), and Random Forest(RF) were employed to predict the water quality class. Experiments were conducted using the real dataset containing the details from various places in Tamil Nadu as well as the synthetic dataset generated on the basis of parameters randomly. Based on the performance of five kinds of classifier, it was found out that the Random Forest classifier achieves some improved result compared to other classifiers. From the analysis it shows that machine learning techniques have the good ability to predict the water quality index