In the present scenario where there are no standard procedure/codal provisions for the mix design of ferrosialate geopolymers, this study proposes a new and efficient way to predict the compressive strength of these mortar specimens by generating prediction equations. This work also studies the influence of various parameters on the strength gain of ferrosialate geopolymers and compares that with conventional fly ash/sialate based geopolymers. Red mud (RM) and fly ash are used as a precursor for geopolymerization, and Gene Expression Programming (GEP) is used for predictive analytics. Ferrosialate mortar samples displayed 112.4% higher compressive strength than sialate mortar samples with denser microstructure and least unreacted phases in the binder. Ideal RM replacement in the binder is determined as 30% and 35% for the oven and ambient curing conditions, respectively. If RM is used as raw feed for ferrosialate geopolymers, the optimum molarity of NaOH solution is found to be 8 M. Oven curing type has higher possibility of strength gain than ambient curing type, and increased curing time leads to a higher possibility of strength gain in ferrosialate geopolymers. The average r2 values for GEP I and GEP II models are 0.88 and 0.92, respectively. Errors of training and validation sets also suggest that a good fit is obtained between observed and predicted values. This indicates that both the models performed well in predicting the compressive strength. Though this model is generated for ferrosialate geopolymer, a similar procedure can be adapted to any geopolymer for the strength prediction. © 2019 Elsevier Ltd