Exploring the potential of the machine learning techniques in the water quality assessment: a review of applications and performance

In this review, the application of machine learning (ML) algorithms in water environment research is proficiently explored. The quick increase in data size related to the water environment has necessitated the use of ML for data analysis, classification, and forecasting. Unlike classical models, machine learning models excel in solving complex problems. They have been successfully applied to various aspects of water management and treatment systems, such as construction, simulation, evaluation, water pollution surveillance, controlling, water quality amelioration, and watershed environmental security management. The survey specifically focuses on the evaluation of water quality in diverse water environments, including surface water, drinking water, groundwater, sewage, and seawater. Moreover, potential future implementations of machine learning in water environments are proposed. ML facilitates the detection and prediction of water contamination events, as well as the provision of decision support systems for water resource management. Real-time monitoring of water quality, anomaly detection, and prediction of potential contamination events are among the specific applications of machine learning high-lighted. The review covers the advantages and disadvantages of generally used ML algorithms, with a particular emphasis on new ML techniques that surpass classical methods.