Water Quality Analysis and Prediction Using Random Forest and Naive Bayes

Abstract

Water quality prediction and monitoring are critical for public health and environmental safety. The proposed research will attempt to conclude a machine learning approach to predict water quality mainly based on parameters like pH, turbidity, temperature, and portability. Two main approaches - the Random Forest Classifier and the Naive Bayes Classifier models - classify water quality into predefined categories, such as "Good" or "Bad," using historical data. The dataset is preprocessed by a correlation matrix to handle missing values and visualize parameter relationships. Following data splitting into training and testing sets, the respective models are trained and evaluated, achieving a very high accuracy in predicting water quality. For the Random Forest model, feature importance analysis is conducted to explain which parameter has contributed to the classification showing major influencing factors on the water quality. The Naive Bayes model provides a probabilistic framework for classification, giving insights into the likelihood of water quality categories based on input parameters. This model could be used for real-time realizations of rapid quality assessments, and environmental agencies, researchers, and the public could get safe water standards. This study shows machine learning's capability to support environmental health initiatives through proactively monitoring water quality.