Machine learning-based prediction of heavy metal immobilization rate in the solidification/stabilization of municipal solid waste incineration fly ash (MSWIFA) by geopolymers.

Geopolymer is an environmentally friendly solidification/stabilization (S/S) binder, exhibiting significant potential for immobilizing heavy metals in municipal solid waste incineration fly ash (MSWIFA). However, due to the diversity in geopolymer raw materials and heavy metal properties, predicting the heavy metal immobilization rate proves to be challenging. In order to enhance the application of geopolymers in immobilizing heavy metals in MSWIFA, a universal method is required to predict the heavy metal immobilization rate. Therefore, this study employs machine learning to predict the heavy metal immobilization rate in S/S of MSWIFA by geopolymers. A gradient boosting regression (GB) model with superior performance (R2 = 0.9214) was obtained, and a graphical user interface (GUI) software was developed to facilitate the convenient accessibility of researchers. The feature categories influencing heavy metal immobilization rate are ranked in order of importance as heavy metal properties > geopolymer raw material properties > curing conditions > alkali activator properties. This study facilitates the rapid prediction, improvement, and optimization of heavy metal immobilization in S/S of MSWIFA by geopolymers, and also provides a theoretical basis for the resource utilization of industrial solid waste, contributing to the environmental protection.