Performance of Artificial Intelligence Models in Analysis and Prediction of Water Potability

Water is a prime necessity for the survival and sustenance of all living beings. Thus, it is very important to maintain a water quality balance. Otherwise, it would seriously damage the health of humans and severely affect the ecological balance among other species. Water quality is an important factor to consider, whether for ecosystem needs or contamination levels that directly impact health, hygiene, food, and the economy. In this study, we have utilized Artificial Intelligence as an efficient technique to predict water quality without resorting to the traditional analysis methods of water monitoring. We have developed a convolutional neural network as a deep learning algorithm and applied five kinds of machine learning algorithms: Backpropagation Neural Network, Random Forest, Decision Jungle, Naive Bayes, Logistic Regression, and Support Vector Machine. The reported results of the classification process showed that the CNN has superior to the other machine learning algorithms in predicting the water potability based on the eight adopted parameters of water properties. CNN has achieved nearly 97% accuracy in the prediction process, while RF, the best one in the utilized machine learning algorithms, has recorded 81%. In addition, CNN has also proven its capability to reduce classification processing time.