The artificial neural network technology simulates the neural network structure of the cerebral cortex by establishing artificial neurons and sensors to solve nonlinear engineering problems, and establishes the prediction model for solving such problems. Rock slope stability is a hot research topic in the field of geotechnical engineering, especially for those slopes affected by the cyclic changes of water level. This paper establishes a slope stability prediction model based on the radial basis function neural network (RBFNN) of the cerebral cortex and introduces the genetic algorithm to eliminate the drawbacks of the cerebral cortex RBFNN, that is, slow convergence and local optimisation. The cerebral cortex RBFNN prediction model can predict the slope safety factor, with the relative error controlled between -5.16% and 6.02%. According to the cerebral cortex RBFNN prediction results, as the water level in the tailing pond changes cyclically, the rock mechanics parameters of the slope gradually weaken, which leads to the continuous decrease of the slope safety factor. The results can provide reference for the extension and application of the cerebral cortex RBFNN in the engineering field, and serve as guidance for the specific project safety maintenance

RBFNN, Genetic Algorithm, Slope Stability, Safety Factor, Water Level Changes