Predicting Water Safety: Harnessing the Power of Simple Machine Learning Algorithms

One of the most valuable resources that is necessary for all life is water. Water pollution lowers the quality of the water, which affects the health of marine life and, consequently, that of humans who utilize it. Because of this, it's imperative to monitor water quality and guarantee the survival of marine life. This research will use artificial intelligence and machine learning techniques to evaluate the efficacy of several models in forecasting water safety. Orange tools are used to apply several prediction models, including Decision Tree, Gradient Boosting, Naive Bayes, Neural Network, SGD, kNN, and CN2 rule inducer. The review employs publicly available secondary data from Keggle, ensuring transparency and accessibility in the research methodology. The results demonstrate that the GB, Neural Network, and Tree achieved superior accuracy ratings of 0.992, 0.998, 0.997, and 0.997, respectively, with remarkable AUC values of 1.000, 1.000, and 0.969. In addition, their recall and accuracy scores are 0.998, 0.998, 0.998, and 0.997, confirming their outstanding performance in predicting water safety.