Modern approaches to greywater pollutant removal techniques

Abstract

This study examined the use of natural and modified zeolite for contaminant adsorption from water and applied machine learning to predict adsorption capacity. A dataset from published studies included parameters such as concentration, contact time, pH, temperature, and zeolite properties. Langmuir and Freundlich models were used to analyze adsorption behavior. Machine learning models including Random Forest and ANN showed high prediction accuracy. Surface area, particle size, concentration, and contact time strongly affected adsorption performance. The study confirmed that combining adsorption modeling with machine learning is effective for wastewater and greywater treatment.