<p>Slope stability evaluation in Geotechnical Engineering is of paramount importance, with direct implications on the environment, public safety through infrastructure and hazard mitigation. The goals of this research were to build deep machine learning models for the purpose of identifying the key parameters that will determine slope stability, and to provide insight into which geo-technical properties have the highest impact on slope stability (SS). After many correlations were tested, it was clear to see that the primary factors affecting slope stability were cohesion, slope height, unit weight, and friction angle. The use of a larger set of slope data allowed us to develop an algorithm to classify slope stability through the implementation of linear regression (LR), convolutional neural networks (CNNs), and long short-term memory networks (LSTMs). The Binary Graylag Goose model (bGGO) used with LSTM was found to have higher sensitivity, predictive value and F1 score than a standard LSTM model. It proved to be a dependable model for intricate predicting tasks in real-world applications by demonstrating strong generalizability and attaining 95% accuracy on a sizable 2160-sample dataset using the bGGO feature selection approach. This study provides useful guidance for geotechnical engineering practices and highlights the value of using feature selection and advanced machine learning algorithm methods for enhancing classification precision for slope stability assessments.</p>