A Comparative Study of Various Classification Techniques to Determine Water Quality

Classification and monitoring the water quality is one of the important aspects which has attracted a lot of attention in the recent years. This work focuses on determining the water quality using different classification techniques such as Decision Tree (DT), K-nearest neighbour (KNN) and Support Vector Machine (SVM) on the ground water samples of Madhya Pradesh, India. The water samples of all 51 districts of Madhya Pradesh which were subjected to chemical analysis were collected. The water samples have been classified (good, average and bad quality) based on the mineral content present in the samples. A comparative study of classification techniques was done based on confusion matrix, accuracy of classification and Receiver Operating Characteristic (ROC). The classification is done based on the electrical conductivity levels. The results suggest that SVM is a better classification model than KNN and DT models on the basis of performance measure.