Slope stability is a crucial aspect of geotechnical engineering, particularly for landfills where municipal solid waste (MSW) layers are subjected to both static and seismic forces. This study represents the first application of hybrid metaheuristic-neural models to the Barmshour Landfill, introducing an innovative predictive framework capable of guiding real-world design, stability evaluation, and decision-making processes in waste management engineering. Four hybrid models-BBO-MLP, MVO-MLP, VS-MLP, and BSA-MLP-were developed and evaluated using real data from the Barmshour Landfill in Shiraz, Iran. The MVO-MLP model achieved the best performance, with coefficient of determination (R2) values of 0.899 (training) and 0.898 (testing), and corresponding RMSEs of 77.60 and 89.44. The results demonstrate that hybrid metaheuristic-neural models can capture complex slope behaviors more effectively than traditional approaches. The primary advancement of this research lies in its systematic comparison of multiple hybrid algorithms and their demonstration of robustness under variable conditions. Practically, the proposed framework provides engineers with a more reliable and adaptive tool for assessing landfill stability and managing geotechnical risks. These findings highlight the growing potential of intelligent hybrid systems to support safer and more data-driven waste management infrastructure.
Mokhtari et al. (Wed,) studied this question.