ABSTRACT This research introduces and validates advanced machine learning models designed to predict the probability of liquefaction failure (pf) in alluvial soil deposits. Three optimisation algorithms namely Northern Goshawk Optimization (NGO), Jellyfish Search Optimizer (JSO), and Horse Herd Optimization Algorithm (HHO) coupled with Adaptive Neuro Fuzzy inference system (AFS) has been employed in the present research. Among the models tested, the AFS-HHO model exhibited superior predictive ability, with R2 = 0.93 and RMSE = 0.06 during the learning stage, and R2 = 0.89 and RMSE = 0.07 during the testing stage. This highlights the AFS-HHO model's efficiency in accurately predicting pf using only corrected SPT-N value i.e. (N1)60 and cyclic stress ratio (CSR). The study also emphasises the importance of the (N1)60 value in influencing the probabilistic assessment of liquefaction failure, and proposes a novel liquefaction probability assessment chart as a novel and reliable tool for estimating liquefaction failure of alluvial soil deposits. Considering the overall analysis, the proposed models offer a novel tool for geotechnical engineers to estimate the probability of liquefaction failure thereby holding substantial implications in the field of probabilistic evaluation in liquefaction studies.