The advancement of geopolymer concrete technology presents a sustainable solution to address the challenges associated with the significant carbon footprint of the construction industry. This research uses machine learning methods to predict the compressive strength of fly ash-based geopolymer concrete using a wide range of experimental results. Furthermore, the effect of different features, including the components of fly ash-based geopolymer concrete, alkaline activator, and other additives, was investigated utilizing Shapley values technique. The results show that both XGBoost and linear regression models can predict the compressive strength of the studied dataset with acceptable accuracy. In XGBoost approach, R2 values for the train and test data were obtained 0.98 and 0.86, respectively, which leads to more accurate results than linear regression method. Additionally, based on the results, MgO is regarded as the most influential factor in compressive strength compared to other investigated features. Furthermore, the concentration of the alkaline activator, NaOH, positively impacts the target value.