A MACHINE LEARNING-BASED MODELING APPROACH FOR DYE REMOVAL USING MODIFIED NATURAL ADSORBENTS

This repository contains the dataset and accompanying scripts used in the study titled "A Machine Learning-Based Modeling Approach for Dye Removal Using Modified Natural Adsorbents" published in J. Chem. Inf. Model. (2025, 65(16), 8486–8496). DOI: 10.1021/acs.jcim.5c01016. The dataset includes adsorption capacity (qₑ) and removal percentage (%) values obtained by removing methylene blue (MB) dye from wastewater using various adsorbent types. These adsorbents were modified by incorporating levulinic acid (LA) into almond, walnut, and apricot kernel powders. The dataset contains experimental parameters including pH, adsorbent dose, initial dye concentration, temperature, and contact time. The primary output variables for modeling are the adsorption capacity (qₑ) and removal percentage (%). This research did not receive external funding. The dataset was derived from the peer-reviewed publication by Süheyla Kocaman (2020), "Removal of methylene blue dye from aqueous solutions by adsorption on levulinic acid-modified natural shells", published in Environmental Technology (22, 885–895). DOI: 10.1080/15226514.2020.1736512. The original data has been reused for machine learning-based modeling purposes with appropriate attribution.